🤖 AIOverlayAgent

A free, open-source AI desktop overlay for developers, students, and professionals.

Instantly ask AI about anything on your screen using screenshots, global hotkeys, and your favorite AI models—without monthly subscriptions or vendor lock-in.

---

🌟 Why AIOverlayAgent?

Most desktop AI assistants require subscriptions, lock you into a single provider, or hide advanced features behind a paywall.

AIOverlayAgent is different.

• 🆓 100% Free & Open Source (MIT)
• 🔑 Bring your own API key
• 🤖 Works with multiple AI providers and models
• 💰 No subscriptions or recurring fees
• 🔒 Your API keys stay under your control
• ⚡ Lightweight, fast, and keyboard-first
• ❤️ Built by developers, for developers

Whether you're solving coding problems, debugging applications, studying, writing documentation, or using AI during meetings, AIOverlayAgent keeps AI assistance just one shortcut away.

---

🖼️ Screenshots

🌐 Landing Page

🖥️ App UI

💻 LeetCode Answer Example

---

✨ Features

• 🤖 AI desktop overlay that stays above other windows
• 📸 Screenshot-based AI assistance with multimodal models
• ⌨️ Powerful global keyboard shortcuts
• 🧠 Support for OpenRouter, OpenAI, Anthropic, and Gemini
• 🎨 AI-powered adaptive themes
• 📝 Beautiful Markdown rendering
• 📤 Export conversations to Markdown
• 👆 Click-through mode
• 👻 Capture exclusion & overlay utilities
• 💾 Local chat history
• 🔒 No account required
• ❤️ Fully open source

---

🚀 Perfect For

• 👨‍💻 Software Developers
• 🎓 Students
• 💼 Professionals
• 📚 Learning & Research
• 🧩 Competitive Programming
• 💻 LeetCode & DSA Practice
• 🐞 Debugging Code
• 🤖 AI-Assisted Workflows

---

❤️ Why Open Source?

AIOverlayAgent is built in the open so everyone can:

• 🔍 Inspect the source code
• 🛠️ Customize every feature
• 🚀 Build their own version
• 🐛 Report bugs
• 💡 Suggest improvements
• 🤝 Submit pull requests

Just an open-source AI desktop assistant that anyone can use.

---

⭐ Support the Project

If AIOverlayAgent helps you, you can support the project by:

• ⭐ Starring the repository
• 🍴 Forking the project
• 🐞 Reporting bugs
• 💡 Suggesting new features
• 🤝 Opening pull requests
• 📢 Sharing it with friends and fellow developers

Every star and contribution helps the project grow.

---

🐞 Issues & Feature Requests

Found a bug? Have an idea?

• 🐞 Open an Issue for bugs or unexpected behavior.
• 💡 Use Discussions for feature requests, questions, and ideas.
• 🚀 Pull requests are always welcome!

Please include reproduction steps, screenshots (if applicable), and your Windows version when reporting issues.

---

🚧 Roadmap

• ✅ Desktop AI Overlay
• ✅ Screenshot AI
• ✅ Multiple AI Providers
• ✅ Markdown Export
• ✅ AI Theme Adaptation
• 🚧 Voice Conversations
• 🚧 Real-time Meeting Assistant
• 🚧 Plugin System
• 🚧 OCR Improvements
• 🚧 Cross-platform Support

Community suggestions help decide what gets built next.

📋 Requirements

• 🪟 Windows 10 build 19041 or newer, or Windows 11.
• 🐍 Python 3.10 or newer for development.
• 🔑 An API key for at least one provider:
  • OPENROUTER_API_KEY
  • OPENAI_API_KEY
  • ANTHROPIC_API_KEY
  • GEMINI_API_KEY

🚀 Installation

git clone https://github.com/suryanshvermaa/personal-agent.git
cd personal-agent
py -3 -m venv .venv
.venv\Scripts\activate
pip install -r requirements.txt

Copy the environment template and add your provider key:

copy .env.example .env

Windows environment variables take priority over .env.

🛠️ Development Setup

Run the app from the repository root:

python main.py

The app starts directly. No activation step is required.

Useful files:

• config.ini: provider, model, UI, capture, and hotkey defaults.
• app_config.ini: app name, executable name, publisher, and version.
• USER_GUIDE.html: single-file browser documentation with copy buttons.
• src/: application code.
• prompts/ and src/prompts/: prompt documentation and runtime prompt profiles.

📦 Build Instructions

Install build-only tools:

pip install -r requirements_build.txt

Build the standalone executable:

build\build_exe.bat

Build the Windows installer after the executable is created:

build\build_installer.bat

The installer build requires Inno Setup 6.

📖 Usage Guide

For a complete end-user guide, see USER_GUIDE.html.

1. 🚀 Start the app with python main.py or the packaged executable.
2. 🔄 Select a model from the header dropdown.
3. 💬 Type a question and send it.
4. 📸 Use screen capture when the model needs visual context.
5. 📤 Export the conversation to Markdown when needed.

Default hotkeys are configured in config.ini:

Hotkey	Action
Ctrl+Shift+Space	Show or hide overlay
Ctrl+Shift+S	Capture screen
Ctrl+Shift+Enter	Send prompt
Ctrl+Shift+E	Export chat
Ctrl+Shift+Alt+T	Toggle click-through

🔒 Privacy

Screenshots and prompts are sent to the AI provider you configure. Avoid capturing sensitive information unless you are comfortable sending it to that provider.

🤝 Contributing

Contributions are welcome. Read CONTRIBUTING.md before opening a pull request, and follow the CODE_OF_CONDUCT.md.

❓ FAQ

Does this project require activation?

No. All licensing and activation code has been removed.

Which provider should I use?

OpenRouter is convenient if you want one key for multiple model vendors. Direct OpenAI, Anthropic, and Gemini integrations are also available.

Where is user data stored?

By default, user data is stored in %APPDATA%\AIOverlayAgent\.

Can I build an installer?

Yes. Use the build commands above and install Inno Setup 6 before running build\build_installer.bat.

Does capture exclusion work everywhere?

No capture-exclusion approach is universal. Test with your target recording or screen-sharing tool before relying on it.

📄 License

This project is licensed under the MIT License. See LICENSE.