Azure Terminal Copilot

What is Azure Terminal Copilot?

Azure Terminal Copilot is a Python-based Azure CLI assistant that provides natural language processing capabilities for Azure commands. It leverages an Azure MCP Server and a local Ollama model to convert human queries into executable Azure CLI commands.

How to use Azure Terminal Copilot?

Clone the repository, create a Python virtual environment with uv, install dependencies, run Ollama and Azure MCP Server locally, configure environment variables in .env, then execute python main.py from the terminal. Provide natural language queries such as "list all my resource groups" to see the corresponding Azure CLI command executed.

Key features of Azure Terminal Copilot

• Natural language to Azure CLI command conversion
• Local execution using Ollama and Azure MCP Server
• Supports any model available through Ollama
• Configurable system prompt for improved responses
• Runs entirely on the user's machine with local dependencies

Use cases of Azure Terminal Copilot

• Quickly listing Azure resources using natural language
• Executing Azure CLI commands without memorizing syntax
• Testing and comparing different language models for command generation
• Learning Azure CLI by observing generated commands from queries
• Automating routine Azure management tasks via conversational input

FAQ from Azure Terminal Copilot

What dependencies does Azure Terminal Copilot require?

Python 3.11+, Azure CLI installed and configured, Ollama with a downloaded model, uv for package management, and a running Azure MCP server.

How does Azure Terminal Copilot authenticate with Azure?

It uses the Azure CLI login; Azure MCP Server uses that same authentication. Ensure you are logged into Azure CLI before running the tool.

Where does the language model run?

The model runs locally via Ollama. You must have Ollama installed and running, and provide its local address in the .env configuration file.

Can I use a different language model?

Yes. In the .env file you can specify any model you have downloaded locally through Ollama. The README suggests trying different models to see which performs better.

Where are queries and data processed?

All processing happens locally on your machine. The tool communicates with locally running Ollama and Azure MCP Server; no cloud service is involved in the conversation.