✍ Proofread — Context Menu Proofreader

A Chrome extension that lets you right-click on selected text and proofread it in different styles using AI — works out-of-the-box with Chrome's Built-in AI (Gemini Nano), no API key needed.

Features

• Zero-config by default — uses Chrome Built-in AI (Gemini Nano), no sign-up or API keys
• Right-click context menu with 8 proofread styles:
  • 📝 Formal — professional, polished tone
  • 😊 Friendly — warm, conversational tone
  • 📧 Email — professional email format
  • 😂 Humorous — witty rewrite
  • ✅ Grammar Only — fix errors, keep original tone
  • 💡 Simplify — plain, clear language
  • 🇬🇧 Translate → English + Proofread
  • 🇫🇷 Translate → French + Proofread
• Floating popup with proofread result, draggable header
• Copy result to clipboard
• Replace selected text in editable fields (textarea, input, contenteditable)
• Try Again button to regenerate with a different variation
• Auto-detects input language and proofreads in the same language (except translate styles)
• Dark/light theme follows system preference

AI Providers (Tiered)

The extension uses a tiered approach so non-technical users can get started immediately:

Tier	Provider	Auth	Default Model	Notes
Default	Chrome Built-in AI	None	Gemini Nano	Zero-config, runs on-device, Chrome 138+
Upgrade	Google Gemini	Sign in with Google	gemini-2.0-flash	Better quality, cloud-based
Advanced	OpenAI	API key	gpt-4o	Hidden behind "Advanced" toggle
Advanced	xAI / Grok	API key	grok-3	Hidden behind "Advanced" toggle
Advanced	Ollama (local)	None	llama3	Hidden behind "Advanced" toggle

Installation

1. Clone or download this repository
2. Open Chrome and go to chrome://extensions/
3. Enable Developer mode (top-right toggle)
4. Click Load unpacked and select the proofread-extension/ folder
5. The ✍ icon appears in your toolbar

Configuration

Right-click the extension icon → Options (or click the ✍ icon)

Chrome Built-in AI (Default)

Works immediately with no setup on Chrome 138+. The options page shows the current status:

• Ready — model is available on-device
• Downloading — model is being downloaded (first use)
• Unavailable — browser/device not supported; switch to a cloud provider

Google Gemini (Upgrade)

1. Select OpenAI (GPT-4o) as provider
2. Paste your API key from platform.openai.com/api-keys
3. Click Save Settings

Google Gemini (Upgrade)

1. Click ⚙️ Advanced Options to expand
2. Select Google Gemini as provider
3. Click Sign in with Google — no API key needed
4. Click Save Settings

Note: For the Google OAuth to work, you need a Google Cloud project with an OAuth 2.0 client ID (type: Chrome Extension). Set the client ID in manifest.json → oauth2.client_id. See Google's guide.

OpenAI (Advanced)

1. Click ⚙️ Advanced Options to expand
2. Select OpenAI (GPT-4o) as provider
3. Paste your API key from platform.openai.com/api-keys
4. Click Save Settings

xAI / Grok (Advanced)

1. Click ⚙️ Advanced Options to expand
2. Select xAI (Grok) as provider
3. Paste your API key from console.x.ai
4. Click Save Settings

Ollama / Local (Advanced)

1. Install Ollama and pull a model: ollama pull llama3
2. Click ⚙️ Advanced Options to expand
3. Select Ollama (Local) as provider
4. Set the base URL (default: http://localhost:11434)
5. Click Save Settings

Usage

1. Select text on any webpage
2. Right-click → ✍ Proofread → choose a style
3. A floating popup appears with the proofread result
4. Actions:
   • 📋 Copy — copy result to clipboard
   • ✏️ Replace — replace selected text (only in editable fields)
   • 🔄 Try Again — generate a new variation
   • ✕ or click outside — close the popup
   • Drag the header to reposition the popup

File Structure

proofread-extension/
├── manifest.json              # Extension manifest (MV3)
├── background/
│   └── service-worker.js      # Context menu + LLM orchestration
├── content/
│   ├── content.js             # Popup overlay rendering
│   └── content.css            # Popup styles
├── lib/
│   ├── llm-providers.js       # LLM provider abstraction + prompts
│   └── google-auth.js         # Google OAuth2 helper
├── options/
│   ├── options.html           # Settings page
│   └── options.js             # Settings logic
├── icons/
│   ├── icon16.png
│   ├── icon48.png
│   └── icon128.png
└── README.md

Security

• API keys stored in chrome.storage.sync (encrypted at rest by Chrome)
• Chrome Built-in AI runs entirely on-device — text never leaves the browser
• LLM responses rendered as plain text (textContent), never innerHTML — XSS-safe
• No eval() or dynamic code execution
• Host permissions scoped to only the required API domains
• Google OAuth tokens managed by Chrome's built-in identity API