Microsoft has launched the Azure Skills Plugin, a comprehensive toolkit designed to transform AI coding assistants into active cloud deployment agents.
The plugin utilizes a three-layer architecture consisting of specialized skills for strategic reasoning, an Azure MCP Server for executing live operations, and a Foundry MCP Server for AI model management. By moving beyond simple code generation, the plugin allows tools like GitHub Copilot and Claude Code to perform real-world tasks such as infrastructure provisioning, cost optimization, and diagnostic troubleshooting.
But let’s get our terminology straight. What is a Plugin and what makes it different from Skills?
A Plugin is a complete, deployable package that bundles both instructional guidance and execution capabilities into a single installation. For example, the Azure Skills Plugin packages together curated “Skills” (the guidance) and “MCP Servers” (the tools) so that AI coding agents can both reason about and actively execute cloud workflows. The plugin’s main purpose is to ensure that the guidance layer and the execution layer stay aligned and can be easily installed across different platforms like GitHub Copilot or Claude Code.
Skills, on the other hand, are just one component contained inside the plugin. They act as the “brain” or guidance layer for the AI. Skills are plain text, version-controlled markdown files (like SKILL.md) that encode domain expertise, reusable workflows, decision trees, and guardrails. They teach the AI how and when to perform tasks, but they do not execute the code or operations themselves. The main difference lies in that a Skill provides the instructions, expertise, and workflows (the “guide”), while the Plugin bundles these skills alongside the MCP Servers; Skills guide, while MCP executes.
The Azure Skills Plugin comprises approximately 20 to 21 curated skills that provide decision trees, workflows, and guardrails for various Azure scenarios, broadly categorized into the following areas:
Build and deploy
Includes azure-prepare (for analyzing projects and generating infrastructure files), azure-validate (for running pre-deployment checks), and azure-deploy (for automating deployment pipelines).
Troubleshoot and operate
Includes azure-diagnostics (for troubleshooting production issues using logs and metrics), azure-observability, and azure-compliance.
Optimize and design
Includes azure-cost-optimization (for identifying potential cloud cost savings), azure-compute, and azure-resource-visualizer.
Data, AI, and platform services
Includes skills for working across different domains, such as azure-ai (which handles AI Search, Speech, OpenAI, and Document Intelligence), azure-aigateway, azure-storage, azure-kusto, azure-rbac, azure-cloud-migrate, entra-app-registration, and microsoft-foundry
Here are some examples of the 19+ curated skills included in the Azure Skills Plugin:
azure-prepare
Analyzes your project and generates necessary infrastructure code, Dockerfiles, and azure.yaml configuration files.
azure-validate
Runs pre-flight checks before deployment to prevent you from wasting time on a failed deployment.
azure-deploy
Orchestrates and automates the actual deployment pipeline using the Azure Developer CLI (azd).
azure-diagnostics
Troubleshoots real production failures and issues by utilizing logs, metrics, and KQL.
azure-cost-optimization
Identifies waste across your cloud resources and provides concrete savings recommendations.
azure-ai
Provides guidance for working with Azure AI services, handling tasks related to AI Search (full-text, vector, and hybrid search), Speech (speech-to-text and text-to-speech), OpenAI, and Document Intelligence .
Other skills include azure-observability, azure-compliance, azure-compute, and azure-resource-visualizer, azure-aigateway, azure-storage, azure-kusto, azure-rbac, azure-cloud-migrate, entra-app-registration, and microsoft-foundry.
Installation is streamlined across various environments including VS Code, IntelliJ, and command-line interfaces to meet developers where they already work. For instance to install the Azure Skills Plugin for VS Code, follow these simple steps:
1. Check prerequisites. Ensure you have VS Code and the Git CLI installed on your machine. If Git is not installed, the skills extension will not function properly.
2. Install the extension. Go to the Extensions panel in VS Code and search for the Azure MCP extension (Extension ID: ms-azuretools.vscode-azure-mcp-server), then install it. This will automatically install a companion extension called GitHub Copilot for Azure, which seamlessly configures the Azure MCP Server, the Foundry MCP Server, and the full Azure skills layer for you.
3. Verify the setup
Open Copilot Chat using Ctrl+Shift+I (Windows/Linux) or Cmd+Shift+I (Mac). Make sure you are using Agent mode instead of Ask or Edit mode. Open the Command Palette (Ctrl+Shift+P / Cmd+Shift+P), search for “MCP”, and confirm that the MCP servers are listed and running.
4. Run a smoke test to prove the installation is working and connected to your account, type this prompt into Copilot Chat: “List my Azure resource groups”. If successful, Copilot will call the live tool (mcp_azure_mcp_group_list) and return your actual resource groups.
Ultimately, the Azure Skills Plugin and its curated skills are useful for Azure professionals because they transform AI coding assistants from passive advisors into active, agentic systems capable of executing real cloud operations.
More Information
Announcing the Azure Skills Plugin
Related Articles
Integrate Azure and Spring with Spring Cloud Azure
To be informed about new articles on I Programmer, sign up for our weekly newsletter, subscribe to the RSS feed and follow us on Facebook or Linkedin.
Comments
Make a Comment or View Existing Comments Using Disqus
or email your comment to: comments@i-programmer.info