Getting Started

Prerequisites

1. Python 3.10+
2. LLM CLI Tool (One of the following):
   • Gemini CLI (Recommended): npm install -g @google/gemini-cli@latest
     • Tip: Enable "Gemini Preview" in settings to access Pro 3/Flash 3 models.
   • Qwen CLI: npm install -g @qwen/cli
   • OpenCode AI: npm install -g opencode-ai
   • Claude Code: npm install -g @anthropic-ai/claude-code

Installation

OS Support

• ✅ Linux & macOS: Fully supported (native).
• ⚠️ Windows: Highly recommended to use WSL2.

Steps

1. Clone the repository:

   git clone https://github.com/yourusername/AgentCommander.git
   cd AgentCommander
   

2. Install dependencies:

   pip install -r requirements.txt
   

3. Start the UI Server:

   bash run_ui.sh
   

   This will start the web server at http://localhost:8080.

Next Steps

Now that the server is running, you have two paths:

1. Initialize a New Project: Use the Auto-Setup Wizard to scaffold your environment.
   • 👉 See "Project Setup" > "Auto-Setup Wizard".
2. Run a Demo: Try the included Diabetes example to see the agent in action immediately.
   • 👉 See "Examples & Guides" > "Diabetes Example".

For advanced configuration (e.g., enabling Gemini Pro), check "Reference & Tips" > "CLI Guide".