Podman AI Assistant

Podman AI Assistant is a command-line tool that helps you manage Podman containers using plain English. Just type what you want to do—like "list all containers"—and it turns that into a real Podman command for you. It's powered by a custom Ollama language model tuned specifically for Podman, making it both smart and practical.

Think of it as an open-source alternative to Docker's Gordon—but for Podman users. It’s designed to be simple to set up and safe to use, with built-in checks for potentially risky commands like rm or stop.

This is just the beginning—up next is MCP server integration, which will bring even more powerful capabilities to the assistant. Stay tuned, more to come!

🔧 Features

• Translates natural language into valid Podman CLI commands Example: "list all containers" → podman ps -a
• Asks for confirmation before running risky commands
• Uses a custom Ollama model (podman-ai) trained on Podman-specific tasks
• Configuration is clean and YAML-based
• Easy CLI access with the podman-ai alias

Here are some sample responses ;)

!Screenshot 2025-05-11 at 3 43 59 PM

!Screenshot 2025-05-11 at 3 47 15 PM

🚀 Quick Start

1. Clone the repo:

   git clone https://github.com/nalinrajendran/podman-agent.git
   cd podman-agent
   

2. Install dependencies:

   • Ollama

   • Podman

   • Python packages:

     pip install -r requirements.txt
     

   • yq YAML CLI parser:

     # Ubuntu
     sudo apt-get install yq
     # Fedora
     sudo dnf install yq
     # macOS
     brew install yq
     

3. Set up your custom model:

   chmod +x bin/create_modelfile.sh
   bin/create_modelfile.sh
   

4. Set up the CLI alias:

   chmod +x bin/setup_podman_ai.sh
   bin/setup_podman_ai.sh
   source ~/.bashrc  # or ~/.zshrc
   

5. Start using it!

   podman-ai "list all containers"
   

For full setup instructions, see docs/SETUP.md.

---

📜 License

MIT License — see LICENSE for more info.

🤝 Contributing

Open to all kinds of contributions! Feel free to open issues or submit PRs on GitHub.

💬 Support

Check the setup guide or open an issue on GitHub.

---