MCP Azure DevOps Server

mcp-azure-devops

An open-source Model Context Protocol (MCP) server for seamless integration with Azure DevOps.

Mission

To create a robust, open-source Model Context Protocol (MCP) server that provides seamless integration with Azure DevOps. This server will empower AI agents to interact with Azure DevOps projects, managing work items, wikis, and repositories, thereby streamlining development workflows.

License

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

Core Features

The MCP server will expose a set of tools to interact with Azure DevOps, categorized by area.

Implemented Features

Work Item Management (CRUD)

• create_work_item (supports Epic, User Story, Task, Bug, and work item linking)
• get_work_item (by ID)
• update_work_item (by ID, supports work item linking)
• delete_work_item (by ID)
• search_work_items (using WIQL - Work Item Query Language)

Wiki Management (CRUD)

• create_wiki_page
• get_wiki_page (by path)
• update_wiki_page (by path)
• delete_wiki_page (by path)
• list_wiki_pages
• get_wikis
• create_wiki

Repository Management (Read-only)

• list_repositories
• list_files (in a repository)
• get_file_content

Project Scoping

• set_project_context: A special tool to set the active project for subsequent commands.
• clear_project_context: To revert to the organization-level scope.
• get_projects: To list all projects in the organization.

Server Documentation

• list_available_tools: Lists all available tools.
• get_tool_documentation: Gets the documentation for a specific tool.

Planned Features

• Repository Management (Write operations):
  • create_repository
  • create_pull_request
  • manage_branches
• Pipeline Management:
  • trigger_build
  • get_build_status
  • list_pipelines

Getting Started

This guide will walk you through setting up the mcp-azure-devops server.

Prerequisites

• Python 3.10 or higher
• pip and venv for managing Python packages

Installation Steps

1. Clone the Repository:
   git clone https://github.com/xrmghost/mcp-azure-devops.git cd mcp-azure-devops
2. Create and Activate a Virtual Environment: It's highly recommended to use a virtual environment to manage the project's dependencies.
   # For Windows python -m venv .venv .\.venv\Scripts\activate # For macOS/Linux python3 -m venv .venv source .venv/bin/activate
3. Install Dependencies: Install the project and its dependencies in editable mode.
   pip install -e .

Configuration

1. Generate an Azure DevOps Personal Access Token (PAT):
   • Navigate to your Azure DevOps organization.
   • Go to User settings > Personal Access Tokens.
   • Click + New Token.
   • Give your token a name (e.g., mcp-server-token).
   • Select the organization.
   • Set the expiration date.
   • For the scopes, you will need to grant the following permissions at a minimum:
     • Work Items: Read & write
     • Wiki: Read & write
     • Code: Read
   • Click Create and copy the token immediately. You will not be able to see it again.
2. Configure the MCP Server in Cline:
   • Open your cline_mcp_settings.json file.

   • Note: The location of this file can vary. A common location on Windows is C:\Users\<YourUsername>\AppData\Roaming\Code\User\globalStorage\saoudrizwan.claude-dev\settings\cline_mcp_settings.json. If you can't find it, you can search your user's home directory for the file.

   • Add a new entry for the mcp-azure-devops server. The command should be mcp-azure-devops.

   Here is an example configuration. You must use the full, absolute path to the mcp-azure-devops.exe executable created inside your virtual environment.

   { "mcpServers": { "mcp-azure-devops": { "command": "C:\\path\\to\\your\\project\\mcp-azure-devops\\.venv\\Scripts\\mcp-azure-devops.exe", "args": [], "env": { "AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/your-organization", "AZURE_DEVOPS_PAT": "your-personal-access-token" }, "disabled": false, "autoApprove": [] } } }
   • Replace your-organization with your Azure DevOps organization name and your-personal-access-token with the PAT you generated.
3. Restart Cline: Restart your Cline application to load the new MCP server.

Acknowledgements

This project was inspired by the mcp-atlassian server, which provides similar functionality for Jira and Confluence. You can find it here: https://github.com/pashpashpash/mcp-atlassian.