DevOps Automation Project

An automation project for DevOps processes using LLM-based agents. Currently focused on local LLM deployment using Ollama.

Features

• 🤖 LLM Integration for intelligent automation (using Ollama)
• 📨 Notifications support via Telegram and Slack
• 🔄 Asynchronous task processing
• 📝 Logging and monitoring

Requirements

• Python 3.8+
• Ollama for local LLM

Installation

1. Install Ollama:

# For macOS or Linux
curl -fsSL https://ollama.ai/install.sh | sh

2. Clone the repository:

git clone git@github.com:whoekage/devops_agent.git
cd devops_agent

3. Create virtual environment:

python -m venv venv
source venv/bin/activate  # For Linux/MacOS
# or
.\venv\Scripts\activate  # For Windows

4. Install dependencies:

pip install -r requirements.txt

5. Copy the example configuration file and set it up:

cp config.yaml.example config.yaml

6. Edit config.yaml and specify the required parameters:

• Ollama model configuration
• Telegram and Slack tokens
• Other parameters as needed

Running

1. Start Ollama and pull the required model:

# Start Ollama service
ollama serve

# In another terminal, pull the model (e.g., mistral)
ollama pull mistral

2. Activate virtual environment (if not already activated):

source venv/bin/activate  # For Linux/MacOS
# or
.\venv\Scripts\activate  # For Windows

3. Run the main application:

python main.py

Project Structure

.
├── agents/             # Agents for various tasks
│   └── k8s_agent.py   # Kubernetes management agent
├── core/              # Application core
│   ├── llm.py         # LLM integration
│   ├── database.py    # Database operations
│   └── notifications.py # Notification system
├── workers/           # Worker processes
│   ├── llm_worker.py  # LLM task processor
│   └── notification_worker.py # Notification processor
├── utils/             # Utility functions
│   └── logger.py      # Logging
├── config.yaml        # Application configuration
├── main.py           # Application entry point
└── requirements.txt   # Project dependencies

Configuration

Main settings are stored in config.yaml. Configuration example:

llm:
  type: "ollama"
  model: "mistral"
  host: "http://localhost:11434"
  temperature: 0.7
  max_tokens: 2000

notifications:
  telegram:
    token: your_telegram_token
    chat_id: your_chat_id
  slack:
    token: your_slack_token
    channel: #notifications

database:
  path: history.db

Development

To contribute to development:

1. Create a fork of the repository
2. Create a branch for new functionality
3. Make changes and create a pull request

TODO

• Implement core notification functionality:
  • Add Telegram bot commands and interactions
  • Implement Slack app integration and slash commands
  • Add message templates and formatting
  • Implement notification priorities and rules
• Add support for multiple LLM providers (OpenAI, Anthropic)
• Implement model switching functionality
• Add Kubernetes integration
• Create comprehensive test suite
• Add metrics collection and monitoring
• Implement rate limiting and queue management
• Add support for more notification channels (Discord, Email)
• Create web interface for management
• Add documentation for API endpoints
• Implement backup and restore functionality
• Add support for custom agent development
• Create examples and use cases
• Add CI/CD pipeline configuration

License

This project is distributed under the MIT License. See LICENSE file for details.