ai-py

A Python CLI application for interacting with Google Gemini and OpenAI (ChatGPT) APIs used for testing vortex tracing agent at https://github.com/flowerinthenight/vortex-agent.git

---

Features

• Flexible CLI: prompt, context, and chatbot modes
• Supports Gemini (REST or google-generativeai) and OpenAI (ChatGPT)
• Easy switching between backends with flags
• Works on Linux, WSL, and Windows

---

Setup

Linux / WSL

1. Get your Gemini API key from Google AI Studio
2. (Optional) Get your OpenAI API key from OpenAI
3. Set environment variables:

   export GEMINI_API_KEY="your_gemini_api_key_here"
   export OPENAI_API_KEY="your_openai_api_key_here"  # if using ChatGPT
   # To make permanent, add to ~/.bashrc or ~/.zshrc

4. Install dependencies:
   • System-wide (no venv):

     sudo apt update
     sudo apt install python3 python3-pip python3-venv python3-requests

   • Use a virtual environment:

     python3 -m venv .venv
     source .venv/bin/activate
     pip install requests openai 
     pip install -q -U google-genai
     

Windows

1. Get your Gemini API key from Google AI Studio
2. (Optional) Get your OpenAI API key from OpenAI
3. Set environment variables:

   $env:GEMINI_API_KEY="your_gemini_api_key_here"
   [Environment]::SetEnvironmentVariable("GEMINI_API_KEY", "your_gemini_api_key_here", "User")
   $env:OPENAI_API_KEY="your_openai_api_key_here"
   [Environment]::SetEnvironmentVariable("OPENAI_API_KEY", "your_openai_api_key_here", "User")

4. Install dependencies:
   • System-wide (no venv):

     pip install requests openai

   • Or, using a virtual environment:

     python -m venv .venv
     .\.venv\Scripts\Activate.ps1
     pip install requests openai

5. (Optional) For Gemini package backend:

   pip install -q -U google-genai

---

Usage

CLI Flags (main.py)

• --prompt "<your question>" : Send a single prompt (default: Gemini)
• --use-chatgpt : Use OpenAI ChatGPT instead of Gemini
• --use-genai : Use google-generativeai package for Gemini backend
• --use-context : Use context.txt and system instruction for advanced summarization (Gemini only)

Example Commands

• Simple Gemini prompt:

  python3 main.py --prompt "What is the capital of Philippines?"

• Simple ChatGPT prompt:

  python3 main.py --use-chatgpt --prompt "What is the capital of Philippines?"

• Pipe input:

  echo "Explain quantum computing" | python3 main.py

• Use context file (Gemini advanced):

  python3 main.py --use-context

• Use google-generativeai backend:

  python3 main.py --use-genai --prompt "Summarize this text."

Chatbot Mode (chat_cli.py)

• Start Gemini chatbot:

  python3 chat_cli.py

• Start ChatGPT chatbot:

  python3 chat_cli.py --use-chatgpt

• Type your message and press Enter. Type exit or quit to end the session.

Sample Questions

• "Summarize the following meeting notes."
• "What are the latest trends in AI?"
• "Write a Python script that reads a CSV file and plots a chart."
• "How do I set up a virtual environment in Python?"

---

Testing

• Linux:

  chmod +x test_env_wsl.sh
  ./test_env_wsl.sh
  echo "Hello, Gemini!" | python3 main.py
  echo "Hello, OpenAI!" | python3 main.py --use-chatgpt

• Windows:

  python test_env.py
  echo "Hello, Gemini!" | python main.py
  echo "Hello, OpenAI!" | python main.py --use-chatgpt

---

Notes

• If you do not provide --prompt, the script will read from stdin.
• For advanced summarization, use --use-context (requires context.txt).
• For chatbot mode, use chat_cli.py (Gemini by default, ChatGPT with --use-chatgpt).
• All API keys must be set as environment variables.