azure-cli-mcp

Category: cloud-devops-mcp-servers
Tags: mcp, azure, cli, cloud
Source: GitHub - jdubois/azure-cli-mcp

Description

azure-cli-mcp is an MCP (Model Context Protocol) Server that wraps the Azure CLI, providing a prompt-driven interface to manage Azure resources. It enables users to interact with Azure through the CLI, list and configure resources, fix issues, and create new services using local credentials. The tool is designed to be run locally and integrates with tools like Claude Desktop and VS Code.

Features

• Full Azure CLI Access: Can perform any operation available in the Azure CLI.
• Resource Listing: List all resources and check their configuration (e.g., rate limits of Azure OpenAI models).
• Configuration and Security Fixes: Address configuration or security issues such as securing a Blob Storage account.
• Resource Creation: Create new Azure resources (e.g., Container Apps, Container Registries) and connect them using managed identity.
• Prompt Interface: Enhanced prompt to improve CLI usability.
• MCP Protocol Implementation: Acts as an MCP server, compatible with tools that support MCP (e.g., Claude Desktop, VS Code Insiders).
• Local Execution: Runs locally using the user's Azure CLI credentials (recommended for security).
• Multiple Install Options:
  • Java Application: Run as a standalone Java app (requires Java 17+).
  • Docker Container: Run in a containerized environment (requires Azure Service Principal credentials).
• Integration with Developer Tools:
  • Claude Desktop: Add as an MCP server in configuration.
  • VS Code Insiders: Configure via the command palette and use with GitHub Copilot in Agent mode.
• Supports stio and http Transports: stio transport is stable; http transport is experimental and intended for future remote use.

Installation

• Using Java:
  • Install Azure CLI and authenticate (az login).
  • Ensure Java 17+ is installed.
  • Download the latest release JAR from GitHub.
• Using Docker:
  • Create an Azure Service Principal and set the AZURE_CREDENTIALS environment variable.
  • Run the Docker container with the appropriate environment setup.

Security Notes

• Designed for local use; running on remote servers is not recommended in the current implementation due to security risks.
• Users should validate commands generated by LLM integrations.

License

MIT License

Pricing

No pricing information provided; the project is open source under the MIT license.