Kairox is an AI-powered program that lets you instantly capture your screen using hotkeys and send it to an LLM (e.g., OpenAI or Claude) for quick Q&A and problem solving.

Kairox removes the friction of copy-pasting into AI tools.

Instead of:

• taking screenshots manually
• switching tabs
• pasting into ChatGPT/Claude

You can:

• press a hotkey
• capture your screen
• get an answer directly in your terminal

Kairox acts as a real-time assistant. For example, if you're working through a problem or studying for a quiz and get stuck, you can instantly capture your screen and receive contextual help without breaking your workflow.

Hotkey → Screen Capture → Context Builder → LLM API → CLI Output

git clone https://github.com/minhduc29/kairox.git
cd kairox

python3 -m venv .venv
source .venv/bin/activate

python -m venv .venv
.venv\Scripts\activate

pip install -r requirements.txt

Create a .env file in the root directory:

ANTHROPIC_API_KEY=your_claude_key_here

⚠️ Do NOT commit your .env file. Use .env.example as a template.

python3 main.py

• Press your configured hotkey to capture screen and sends to LLM
• Output will be displayed in the terminal

(More modes/hotkeys coming soon)

This project is built for:

• personal use
• experimentation
• learning

The codebase is customized to my needs. That said, it currently only works on macOS. However, it’s structured so others can explore, learn from it, or build on top of it.

Contributions, ideas, and experiments are welcome. Feel free to open issues or PRs.