🚀 IDEPocket - Your PC in Your Pocket

Official Website: idepocket.com

Warning

BETA / DEVELOPMENT ONLY: This project is NOT a stable release. It is currently in active development, experimental, and may contain critical bugs or security flaws. Use at your own risk.

Contribute to IDEPocket! 🚀 We are open to and hope to receive Pull Requests from anyone who wants to continue or complete this project. Your contribution is valuable!

IDEPocket is a revolutionary open-source AI assistant that allows you to control your PC directly from your smartphone. It's not just a remote control: it's an autonomous agent capable of executing complex tasks, writing code, managing the terminal, and automating your workflow wherever you are.

Important

Demo coming soon! We are preparing a video demo to showcase the power of IDEPocket.

---

✨ Key Features

• 🤖 Multi-LLM Support: Choose your favorite brain. Native support for OpenAI (GPT-5.5), Anthropic (Claude 4.7), Groq, Qwen, and Llama.
• 🛡️ Privacy First: Want total control? Use Ollama to run local models. Your data never leaves your home network.
• 🖥️ Native Computer Use: The agent interacts with the desktop like a human would: clicks, types, scrolls, and manages windows.
• 🛡️ Deterministic Security Layer: Powered by YecoAI Security Layer. Heuristic filtering for prompt injections, PII/DLP protection, and execution sandbox guards with < 1.5ms latency.
• 📱 Mobile Experience: Smooth app developed in React Native for immediate and intuitive control.
• 🛠️ Extensible: Modular architecture designed for developers. Add your tools and customize the agent's behavior.

---

🏗️ Project Structure

Module	Description
Backend	FastAPI API, session management, database, and agent orchestration.
Terminal_Agent	The operational "arm" to run on the target PC.
client_gui	The mobile user interface (Android & PWA).
shared	Core logic, communication protocols, and tool definitions.

---

🚀 Quick Start (One-Click)

The fastest way to start IDEPocket (Backend and Agent together) is using our quick start scripts.

1. Clone the repository:

   git clone https://github.com/YecoAI/IDE-Pocket.git
   cd IDE-Pocket

2. Start the environment:

   Windows (PowerShell):

   .\start_local.ps1

   Linux / macOS:

   chmod +x start_local.sh
   ./start_local.sh

Note: On the first run, the script will generate the Backend/.env file. Configure your API Keys inside and run the script again!

---

🛡️ Security & Architecture (IMPORTANT)

Since IDEPocket provides command execution and desktop control capabilities, security is our absolute priority. We implement a Defense-in-Depth strategy:

• YecoAI Security Layer (Heuristic Filter): Every input and output is processed by a deterministic security engine:
  • Asimov Rules Injection: Injects foundational safety guidelines into the LLM system prompt.
  • SafetyModel Validation: Scans LLM outputs for dangerous commands, SQLi, and sensitive data (API keys, etc.) before they reach the user.
  • Execution Sandbox Guard: Intercepts tool calls at runtime to prevent destructive operations (e.g., rm -rf /) and SSRF attacks.
• Localhost Execution: By default, the Backend and Terminal Agent communicate locally. No port is exposed to the public internet.
• Total Privacy with Ollama: You can run LLM models locally using Ollama. No data, screenshots, or code snippets will ever leave your home network.
• JWT Authentication: Remote access via the mobile app is protected by secure JWT tokens.
• Principle of Least Privilege: The Terminal Agent inherits only the permissions of the current user. It is highly recommended not to run the agent as root or Administrator.

---

🛠️ Manual Installation (Docker & More)

Start Backend via Docker

You can easily start the Backend using Docker Compose. Make sure you have configured the Backend/.env file before starting.

docker-compose up -d

1. Backend (Server) - Manual Start

pip install -e ./shared
cd Backend
pip install -r requirements.txt
python main.py

2. Terminal Agent (Target PC)

pip install -e ./shared
cd Terminal_Agent
pip install -r requirements.txt
python main.py

3. Mobile App (Smartphone)

cd client_gui/android
npm install
npx expo start

---

🤝 Contributing

Contributions are what make the open source community such an amazing place. Any help is welcome!

1. Fork the Project
2. Create your Feature Branch (git checkout -b feature/AmazingFeature)
3. Commit your Changes (git commit -m 'Add some AmazingFeature')
4. Push to the Branch (git push origin feature/AmazingFeature)
5. Open a Pull Request

---

📜 License

Distributed under the MIT License. See LICENSE for more information.

---

Developed with ❤️ by HighMark [ Marco N. ] @ YecoAI