GitHub - pavanjava/devops-agent: Agent team for DevOps that solves Kubernetes, Docker, Terraform, and monitoring challenges through intelligent automation.

An AI-powered CLI tool to assist with DevOps troubleshooting, Applications with Kubernetes architecture, log analysis, and infrastructure code generation.

Features

📊 Log Analysis: Analyze log files and get actionable insights
💬 Query Interface: Ask questions about DevOps best practices, Terraform, Kubernetes, etc.
🛠️ Template Generation: Generate infrastructure code templates
🤖 AI-Powered: Leverages multiple LLM providers (OpenAI, Anthropic, Gemini, Ollama, vLLM)
🎯 Flexible Provider Selection: Choose your preferred LLM provider and model dynamically
🔒 Self-Hosted Options: Run privately with Ollama or vLLM
🧠 Reasoning Mode: Enable advanced reasoning capabilities for complex queries
🐛 Debug Mode: Troubleshoot agent behavior with detailed logging
💾 Memory Management: Persistent context using Qdrant vector database
🎨 Interactive Mode: Engage in continuous conversations with the agent
📝 Multiple Output Formats: Export results as text, JSON, or Markdown

Installation

# Clone the repository
git clone https://github.com/yourusername/devops-agent.git
cd devops-agent

# Install in development mode
pip install -e .

# Or install from PyPI (when published)
pip install devops-agent

Configuration

LLM API KEYS

# For OpenAI
export OPENAI_API_KEY=YOUR_API_KEY

# For Anthropic Claude
export ANTHROPIC_API_KEY=YOUR_API_KEY

# For Google Gemini
export GEMINI_API_KEY=YOUR_API_KEY

# For Ollama (self-hosted, typically no API key needed)
export OLLAMA_API_KEY=YOUR_API_KEY  # Optional

# For vLLM (self-hosted)
export VLLM_API_KEY=YOUR_API_KEY

Qdrant Config for Agent Memory

(If not configured fall backs to in-memory vector store)

export QDRANT_URL=YOUR QDRANT URL
export QDRANT_API_KEY=YOUR QDRANT API KEY

Usage

Ask Questions

devops-agent run --query "I need terraform script to spin up Azure blob storage"
devops-agent run --query "How to increase my pod memory and CPU in k8s"

Interactive Mode

devops-agent run --interactive
# or
devops-agent run -i

Advanced Options

Choose Your LLM Provider and Model

# Use OpenAI with a specific model
devops-agent run --provider openai --model gpt-4o --query "your question"

# Use Anthropic Claude
devops-agent run --provider anthropic --model claude-sonnet-4-20250514 --query "your question"

# Use Google Gemini
devops-agent run --provider google --model gemini-2.0-flash-exp --query "your question"

# Use Ollama (self-hosted)
devops-agent run --provider ollama --model llama3 --query "your question"

# Use vLLM (self-hosted)
devops-agent run --provider vllm --model your-model-name --query "your question"

Enable Debug Mode

devops-agent run --query "your question" --debug_mode true

Enable Reasoning Mode

devops-agent run --query "your question" --reasoning_enabled true

Combine Multiple Options

# Interactive mode with specific provider, model, and reasoning
devops-agent run -i --provider anthropic --model claude-sonnet-4-20250514 --reasoning_enabled true

# Query with debug mode and custom output
devops-agent run --query "docker setup for microservices" --provider openai --model gpt-4o --debug_mode true --output result.md --format markdown

CLI Options Reference

`devops-agent run` Options

Option	Type	Description
`--log-file`	Path	Path to log file to analyze
`--provider`	String	LLM provider (openai, anthropic, google, ollama, vllm)
`--model`	String	Model name (e.g., gpt-4o, claude-sonnet-4-20250514, gemini-2.0-flash-exp)
`--query`	String	Query to ask the DevOps agent
`--output`	Path	Output file path for saving results
`--format`	Choice	Output format: text, json, or markdown (default: text)
`--interactive, -i`	Flag	Run in interactive mode for continuous conversation
`--debug_mode`	Boolean	Enable debug mode with detailed logging
`--reasoning_enabled`	Boolean	Enable reasoning mode for complex problem-solving

Provider-Specific Model Examples

OpenAI:

gpt-4o
gpt-5-mini
gpt-5.1

Anthropic:

claude-sonnet-4-20250514
claude-sonnet-4-5-20250929
claude-3-5-sonnet-20241022

Google:

gemini-3-pro
gemini-2.5-pro
gemini-2.5-flash

Ollama (Self-hosted):

granite4:3b
qwen3:8b
cogito:latest
Any model you have pulled locally

vLLM (Self-hosted):

Any model served by your vLLM instance

Development

# Install development dependencies
pip install -e ".[dev]"

# Run tests
pytest

# Format code
black devops_agent/
isort devops_agent/

# Lint
flake8 devops_agent/

Project Structure

devops-agent/
├── devops_agent/          # Main package
│   ├── cli.py            # CLI interface
│   ├── core/             # Core functionality
│   ├── templates/        # Template generators
│   ├── utils/            # Utilities
│   └── prompts/          # LLM prompts
└── docs/                 # Documentation

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

Apache2.0 License - see LICENSE file for details

RoadMap

Implement log analysis with pattern detection
Add Support for MCP to use local file system for quick access
Add support for Human-in-the-Loop for more focused and collaborated work
Support for custom prompt templates
Agent as a Service with privacy first concept

Support

For issues and questions, please open an issue on GitHub.

Special Credits

Built with Agno2.0 framework for multi-agent orchestration
Uses POML for structured prompt engineering
Uses Qdrant for memory management
powered by Claude (Anthropic), GPT (OpenAI) and Gemini (Google)