Zero-config, multi-provider AI assistant to automatically solve and fill Google Forms across Chrome, Firefox, and Edge.
AutoForm AI is a cross-browser extension backed by a high-throughput proxy server that intelligently coordinates Groq (Llama 3), Google Gemini, and NVIDIA NIM with automatic failover, sub-second latency, and zero setup required for end users.
- ⚡ Zero-Config & Instant: Users install the extension and start filling forms immediately — no API keys or setup wizards required.
- 🧠 Multi-Provider AI Engine:
- Groq (
llama-3.3-70b-versatile): Sub-300ms ultra-fast inference (Primary). - Google Gemini (
gemini-2.5-flash-lite): High-accuracy multimodal reasoning (Secondary). - NVIDIA NIM (
meta/llama-3.1-70b-instruct): Enterprise backup fallback.
- Groq (
- 🔄 Auto-Failover & Health Monitoring: If a provider experiences rate limits (
429) or errors, the router automatically fails over to the next provider seamlessly. - 🌐 Universal Cross-Browser Compatibility: Ready for Google Chrome, Mozilla Firefox, and Microsoft Edge on Manifest V3.
- 🎨 Modern Dark UI/UX:
- Redesigned popup with live form scanner and question counter.
- Interactive modal overlay with real percentage progress bar, question preview, and instant "Stop & Cancel" button.
- Floating pill action button directly on Google Forms.
- 🎯 Smart Preferences: Customize answer tone (Accurate, Concise, Detailed) or supply persona context (e.g., "Computer Science student").
- 🛡️ Rate-Limited & Secure: Keys are protected on the backend; users receive generous hourly quotas with client-side rate limit tracking.
AutoForm-AI/
├── manifest.json # Universal Manifest V3 (Chrome, Firefox, Edge)
├── package.json # Root scripts & extension bundler
├── AGENTS.md # Comprehensive AI/LLM developer guide
│
├── server/ # Production Backend Proxy Server
│ ├── package.json # Express, CORS, Dotenv
│ ├── Dockerfile # Production container definition
│ ├── railway.json # 1-click Railway deploy template
│ ├── render.yaml # 1-click Render deploy template
│ ├── .env.example # Server environment template
│ └── src/
│ ├── index.js # Express API server entry point
│ ├── config.js # Multi-provider configuration & key parser
│ ├── providers/ # Provider adapters (Groq, Gemini, NVIDIA)
│ ├── services/ # Multi-provider router & failover engine
│ └── middleware/ # Sliding-window rate limiter & validator
│
├── src/ # Browser Extension
│ ├── background/
│ │ └── background.js # Service worker & backend communication bridge
│ ├── content/
│ │ └── content.js # Google Forms DOM automation & modern UI overlay
│ └── popup/
│ ├── popup.html # Sleek modern popup interface
│ ├── popup.js # Popup controller & form detector
│ └── popup.css # Dark minimalist styling
│
├── docs/
│ ├── ARCHITECTURE.md # Mermaid sequence diagrams & architecture specs
│ └── SELECTORS.md # Google Forms ARIA DOM selector dictionary
│
└── scripts/
├── validate_manifest.js # Manifest & file path validator
└── package_extension.js # Release zip packager
- Navigate to the
server/directory:cd server cp .env.example .env - Open
.envand add your API keys (supports comma-separated keys for multi-key rotation):GROQ_API_KEYS=gsk_your_groq_key_here GEMINI_API_KEYS=AIzaSy_your_gemini_key_here NVIDIA_API_KEYS=nvapi-your_nvidia_key_here
- Start the server:
npm install npm start # Server runs on http://localhost:3000
- Navigate to
chrome://extensions(oredge://extensions). - Toggle on Developer mode in the top right.
- Click Load unpacked and select the root
AutoForm-AIdirectory.
- Navigate to
about:debugging#/runtime/this-firefox. - Click "Load Temporary Add-on...".
- Select
manifest.jsonin the rootAutoForm-AIdirectory.
- Open any Google Form.
- Click the floating "⚡ AI Fill" button on the bottom right (or click "Fill Current Form" from the extension popup).
- Watch AutoForm AI intelligently complete all questions with real-time percentage progress!
- Need to halt? Press the Escape key or click "Stop & Cancel" at any moment.
Deploy the backend to the cloud for free with zero maintenance:
| Platform | Deployment Guide |
|---|---|
| Railway | Connect your GitHub repository → Railway will automatically detect server/railway.json. Set environment variables in the Railway dashboard. |
| Render | Create a new Web Service pointing to server/ with Node environment. Uses server/render.yaml. |
| Docker | docker build -t autoform-ai-server server/ then docker run -p 3000:3000 --env-file server/.env autoform-ai-server |
Once deployed, update the Backend Proxy URL in the extension popup's Advanced Settings (or set DEFAULT_SERVER_URL in src/background/background.js).
- Start backend in watch mode:
npm run server:dev
- Validate extension manifest:
npm run validate
- Package extension for release (
dist/autoform-ai.zip):npm run package
MIT © whoisadheep