DevOps Generator CLI (with Ollama AI)

This is a Python-based CLI tool that leverages Ollama models (like LLaMA3, Mistral, etc.) running locally to auto-generate DevOps infrastructure files such as:

• Dockerfile (with best practices)
• AWS CodeBuild buildspec.yml
• Terraform configurations
• CloudFormation templates
• Ansible playbooks
• Kubernetes YAMLs (Deployment, Service, Ingress, etc.)
• Kyverno security policies
• Helm charts

---

Features

• Uses locally running Ollama models via the ollama Python package
• Prompts for essential input only
• Auto-formats YAML/JSON output
• Supports one task per run (ensures focused results)
• Follows DevOps best practices by default

---

Prerequiste

Installing Ollama

Download and Install Ollama

# For Linux
curl -fsSL https://ollama.com/install.sh | sh

# For MacOS
brew install ollama

Start Ollama Service

ollama serve

Pull Llama3 Model

ollama pull llama3.2:1b

Requirements

• Python 3.7+
• Local Ollama instance running (e.g., ollama run llama3)
• Required Python packages:

pip install -r requirements.txt

requirements.txt ollama PyYAML Usage

python devops_gen.py

Working

You’ll be prompted to choose one of the following:

1. Dockerfile
2. BuildSpec (buildspec.yml)
3. Terraform
4. CloudFormation
5. Ansible Playbook
6. K8s Deployment
7. K8s Service
8. K8s Ingress
9. Kyverno Pod Security
10. Helm Chart Then answer task-specific questions like:

💬 Language of your app (for Dockerfile)

🧪 How the app runs (e.g., Flask with Gunicorn, Express server)

⚙️ Resource limits (for Kubernetes)

🌐 Infrastructure type (e.g., EKS, EC2, RDS)

🌍 Ingress controller (e.g., NGINX, AWS ALB)

📁 Output Generated files are saved directly to the current directory, e.g.:

Dockerfile

terraform.tf

k8s_deployment.yaml

cloudformation.yaml

helm_chart.yaml

Output is also prettified automatically if it's YAML or JSON.

Security

All generation happens locally. No data is sent over the internet — you're using your own locally running Ollama model.