Amadeus_API_CLI

A command-line interface (CLI) tool for interacting with AI models through the Amadeus framework.
It provides a state-driven chat experience with support for configurable prompts, attachments, chat history, and proxy settings.

Features

• Interactive Chat
  Conversational interface powered by the OpenAI API with streaming responses.

• Configurable System Prompt
  Load custom prompts from file or inline to tailor assistant behavior.

• Attachment Support
  Attach images, documents, or text files to enrich conversations.
  Supported formats: .jpg, .png, .pdf, .docx, .txt, .csv, etc. (max 20MB per file).

• Chat History
  Save, load, and clear chat sessions (including attachments and reasoning traces).

• State Machine Design
  The application flow is managed using a finite state machine for clarity and robustness.

• Customization
  Configure API key, model, base URL, proxy, and temperature via CLI or saved configuration file.

Installation

Clone the repository:

git clone https://github.com/starfield17/Amadeus_API_CLI.git
cd Amadeus_API_CLI/V1

Install requirements:

bash install_requirement.sh

(Optional) Add to system path:

bash add_to_sys.sh

Usage

Run the CLI:

python Amadeus_rebuild.py --api-key <YOUR_API_KEY> --model <MODEL_NAME>

Commands

• /clear – Clear chat history and attachments
• /save [filename] – Save chat history to a file
• /load [filename] – Load chat history from a file
• /attach <file> – Attach one or more files
• /detach [index] – Remove attachments
• /files – List current attachments
• /help – Show help and available commands

Keyboard Shortcuts

• Enter: Send message
• Ctrl+D: Send message
• Ctrl+V: Paste text
• Ctrl+Z / Ctrl+Y: Undo/Redo
• Up/Down: Navigate history

Configuration

Configuration is stored in ~/.amadeus_config.toml. Run with --save-config to persist current options.