OpenShift Forward - AI Companion

OpenShift Forward (OSF) is an AI-first application to manage an OpenShift cluster alongside OpenShift Container Platform (OCP) web console. Rather than "shift left" or "shift right", we want to "shift forward"! Note: Support for Kubernetes cluster is limited.

⚠ Beware: This is a prototype only. DO NOT USE IN PRODUCTION!

Log into your cluster and give it prompts. It will create YAMLs for you. You can edit the YAML to your liking and then apply the changes. WARNING: YAML creation is a work in progress!

Demo

Try it live at: https://kubecloud.site

Install and Setup

Environment: This should work on any Linux-based, MacOS, or WSL environment.

Prerequisites:

• OpenAI key

• Binary copy of oc (OpenShift/Kubernetes) or kubectl (Kubernetes only)

1. Install packages

   # Debian environment
   sudo apt update
   sudo apt install python3 python3-pip npm ffmpeg docker.io
   

2. (Optional) Set up virtual environment

   1.1 Use either Python's venv (or Anaconda).

   # venv
   sudo apt install python3.12
   python3.12 -m venv pyenv
   

   1.2 Activate the virtual environment.

   source pyenv/bin/activate
   

3. Set up Docker

   • The user running the Flask app must have access to the docker command without requiring sudo privileges.

   sudo usermod -aG docker user  # replace user with the actual user
   # Log out and log back in for it to take effect.
   

   • Copy the binary oc (OpenShift/Kubernetes) or kubectl (Kubernetes only) into this directory. If you are using kubectl, change the Dockerfile and replace the COPY and RUN lines with COPY kubectl /usr/local/bin/kubectl.

   • Create the Docker container.

   cd docker
   ./run.sh
   docker images  # verify container exists
   

4. Install Python packages

   cd ../backend
   pip install -r requirements.txt
   
   # if running Python 3.13+
   pip install audioop-lts
   

5. Set up OpenAI key and Flask key

   Create a file named ".env" in the backend directory. Replace with your keys. It should contain:

   OPENAI_API_KEY='<your_openai_key>'
   FLASK_SECRET_KEY='<some_random_key>'
   

   FLASK_SECRET_KEY can be any arbitrary value.

6. Set up frontend

   cd ..
   npm create vite@latest frontend -- --template react
     > Choose "Ignore files and continue".
   cd frontend
   npm install
   npm install lucide-react
   
   rm -rf .gitignore README.md eslint.config.js public src/App.css src/assets
   git restore index.html src/App.jsx src/index.css src/main.jsx
   

7. (Optional) Set up Ollama

   • Follow instructions to install Ollama.
   • Get some models using ollama pull <model_library>. In OSF, you should see the models by clicking the Settings icon in the upper right corner.

   Note: In OSF, your LLM selection is not preserved and will be reset on a new session.

Run application

1. (Optional) Switch to your virtual environment

   cd ..
   # if not done already
   source pyenv/bin/activate
   

2. Run app in development mode

   cd backend
   python app.py
   

   In another session,

   cd frontend
   npm run dev
   

   Point your browser at: http://localhost:5173

   Note: If you want to change the port, take a look at frontend/vite.config.js and backend/app.py.

   Happy shifting! 😀