Skip to main content

Azure MCP Agent for secure, compliant resource deployment

Project description

Azure SFI MCP Agent - Installation Guide

Description

Azure SFI MCP Agent is a Model Context Protocol (MCP) server that enables secure, compliant Azure resource deployment directly from VS Code using GitHub Copilot Chat. This agent helps you create SFI compliant Azure resources with automatic compliance orchestration.

Capabilities

Azure Resource Management

  1. List Azure Permissions - View your active RBAC role assignments and access levels
  2. List Azure Resources - Browse resources across subscriptions and resource groups
  3. Create Resource Groups - Create Azure resource groups with project tagging
  4. Create SFI-Compliant Resources - Deploy Azure resources with automatic compliance features:
    • Storage Accounts (ADLS Gen2)
    • Key Vaults
    • Azure OpenAI
    • AI Search
    • AI Foundry
    • Cosmos DB
    • Log Analytics Workspaces
    • Network Security Perimeters (NSP)
    • User Assigned Managed Identity (UAMI)
    • Fabric Capacity
  5. Add Diagnostic Settings - Configure Log Analytics monitoring for resources
  6. NSP Attachment - Configure Network Security Perimeter attachment for supported resources

Azure DevOps Integration

  1. Create DevOps Projects - Set up new Azure DevOps projects with initial repositories
  2. Create DevOps Repositories - Add new Git repositories to existing projects
  3. Create DevOps Branches - Create branches in repositories from base branches
  4. Deploy Pipeline YAML - Deploy pipeline templates (CodeQL, 1ES) to repositories
  5. Create DevOps Pipelines - Create and configure Azure Pipelines from YAML files
  6. List DevOps Projects - View all projects in an organization
  7. List DevOps Repositories - View all repositories in a project

Microsoft Fabric Integration

  1. Create Fabric Workspaces - Create workspaces in Fabric capacities
  2. Attach Workspace to Git - Connect Fabric workspaces to Azure DevOps repositories
  3. List Fabric Permissions - View workspace permissions and access levels

Prerequisites

Before installing the Azure SFI MCP Agent, ensure you have the following installed:

Required Software

  1. Visual Studio Code - Download
  2. PowerShell Core (pwsh) - Download
  3. Azure CLI - Download
  4. Python 3.10+ - Download
  5. uvx - Download
  6. GitHub Copilot Chat Extension - Install from VS Code Marketplace

Azure Requirements

  • Active Azure subscription
  • Appropriate Azure RBAC permissions for resource creation
  • Azure CLI authenticated (az login)
  • Set context for one subscription (az account set --subscription <subscriptionid>)

ADO Requirements

  • Access to Azure DevOps organization
  • Project Collection Admin permissions for creating projects
  • Project Admin permissions for creating repositories, and pipelines
  • Azure CLI authenticated (az login or allow-no-subscriptions)

Fabric Requirements

  • Access to Microsoft Fabric workspaces
  • Appropriate permissions to create and manage workspaces
  • Fabric capacity available for workspace creation
  • ADO Available for GIT integration
  • Azure CLI authenticated (az login or allow-no-subscriptions)

Installation Steps

Step 1: Open GitHub Copilot Chat

  1. Launch Visual Studio Code
  2. Open GitHub Copilot Chat (click the chat icon in the sidebar or press Ctrl+Alt+I)

Step 2: Access MCP Tools Menu

  1. In the Copilot Chat window, click on the 🔧 Tools button
  2. Select "Install MCP Server from PyPI" or similar option

Step 3: Install the Package

  1. When prompted for the package name, enter:
    azuresfimcpagent
    
  2. Select the latest version when prompted
  3. Wait for the installation to complete

Step 4: Configure MCP Settings

Add the following configuration to the mcp.json file:

{
    "mcpServers": {
        "azuresfimcpagent": {
            "type": "stdio",
            "command": "uvx",
            "args": [
                "azuresfimcpagent==1.0.0"
            ]
        }
    }
}

Note: Replace 1.0.0 with the latest version number you installed.

Step 5: Restart VS Code

  1. Close and reopen Visual Studio Code to load the MCP server configuration
  2. Open GitHub Copilot Chat again
  3. Select the MCP Tool installed

Step 6: Verify Installation

In GitHub Copilot Chat, type:

show menu

You should see the available actions menu confirming successful installation.


💡 Usage Examples

Azure Resource Management

List Your Azure Permissions

list my azure permissions

List Azure Resources

list resources in resource-group-name

Create a Resource Group

create resource group named my-rg in eastus for project MyProject

Create a Storage Account

create storage account

Create a Key Vault

create key vault

The agent will interactively prompt you for required parameters and automatically:

  • ✅ Deploy the SFI compliant resources
  • ✅ Configure Log Analytics diagnostic settings
  • ✅ Apply security best practices and compliance controls

Azure DevOps Operations

Create a DevOps Project

create azure devops project named MyProject with repo MainRepo in organization myorg

Create a DevOps Repository

create devops repository named MyRepo in project MyProject

Create a Branch

create branch feature/new-feature from main in MyRepo

Deploy Pipeline YAML

deploy codeql pipeline yaml to MyRepo in pipelines folder

Create a Pipeline

create pipeline named MyPipeline-1ES for MyRepo

List DevOps Projects

list all devops projects in organization myorg

List DevOps Repositories

list all repos in project MyProject

Microsoft Fabric Operations

List Fabric Permissions

list my fabric permissions

Create a Fabric Workspace

create fabric workspace named MyWorkspace in capacity /subscriptions/.../capacities/mycapacity

Attach Workspace to Git

attach fabric workspace to azure devops git

Azure CLI Authentication

Ensure you're logged into Azure CLI:

az login
az account show

PowerShell Core Required

This agent requires PowerShell Core (pwsh), not Windows PowerShell. Verify:

pwsh --version

📄 License

MIT License - see LICENSE file for details

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

azuresfimcpagent_test-2.0.0.tar.gz (41.8 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

azuresfimcpagent_test-2.0.0-py3-none-any.whl (53.5 kB view details)

Uploaded Python 3

File details

Details for the file azuresfimcpagent_test-2.0.0.tar.gz.

File metadata

  • Download URL: azuresfimcpagent_test-2.0.0.tar.gz
  • Upload date:
  • Size: 41.8 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.10.11

File hashes

Hashes for azuresfimcpagent_test-2.0.0.tar.gz
Algorithm Hash digest
SHA256 f492c17d906f42c159f3490558ecc1ee5a90b21e00732c31867096408f5c0607
MD5 16cc0224fefd72f684b6fadc7b6fde67
BLAKE2b-256 ec794bd164233ca9bbe6d40600bcb75dba296a52595134e2ff72fadc6b77aa9e

See more details on using hashes here.

File details

Details for the file azuresfimcpagent_test-2.0.0-py3-none-any.whl.

File metadata

File hashes

Hashes for azuresfimcpagent_test-2.0.0-py3-none-any.whl
Algorithm Hash digest
SHA256 ec23b56dac03b5a61a22cbe46445cf38718d5f2e6eabe161672a8429d969e35e
MD5 0d17711c3388563f535622b02380e659
BLAKE2b-256 c0fb5e73a5a20ad374284b948ad87d722f79dd42201148599394adb331094d4b

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page