AgentCore

A local-first computer operator for Windows.

AgentCore lets you interact with your computer using simple, natural language commands. It can find files, open folders, read files, launch applications, and search the web.

Features

• 🔍 Fast file and folder search using a local index
• ⚡ Live file watching for common user folders
• 📂 Open files and folders
• 📖 Read text-based files directly in the terminal
• 💻 Open files and folders in VS Code
• 🚀 Find and launch installed applications
• 🌐 Search the web using your preferred browser
• 🗣️ Simple natural language commands

Examples

Find mypanpic
Open mypanpic
Read main.py
Open agentcore in VS Code
Open Chrome
Open Chrome and search for committ.in

Installation

Clone the repository

git clone https://github.com/kkj4u2t7/agentcore.git
cd agentcore

Create a virtual environment

python -m venv .venv

Activate it on Windows:

.venv\Scripts\Activate.ps1

Install dependencies

pip install -r requirements.txt

Run AgentCore

python main.py

How It Works

AgentCore uses a local file index to make searching faster than scanning your entire computer every time.

Your instruction
      ↓
Command interpretation
      ↓
File / App / Browser routing
      ↓
Local computer action

The file watcher keeps the index updated when files are added or changed in watched locations.

Project Structure

agentcore/
├── tools/          # Computer interaction tools
├── agent/          # Agent runtime components
├── tests/          # Tests
├── data/           # Local data, not committed to Git
├── main.py         # Application entry point
└── requirements.txt

Roadmap

• Fast indexed file search
• Live file watching
• Open files and folders
• Read files from the terminal
• Launch installed applications
• Browser web search
• More natural language commands
• Command history
• Safer action confirmation
• Cross-platform support
• AI-powered command understanding

Contributing

Contributions, ideas, and bug reports are welcome.

1. Fork the repository
2. Create a feature branch
3. Make your changes
4. Submit a Pull Request

License

This project is licensed under the MIT License.

Author

Created and maintained by Avinash.