Azure DevOps MCP Server
A read-only MCP server connecting AI assistants to Azure DevOps Server (on-premises) for browsing projects, repos, builds, work items, releases, pipelines, and test results.
README
Azure DevOps MCP Server
A READ-ONLY MCP (Model Context Protocol) server that connects AI assistants to Azure DevOps Server (on-premises). Works with Claude Code CLI, VS Code AI extensions (Continue, Cline), Cursor, and any MCP-compatible client.
Features
- Projects: List and explore projects
- Git/Repos: Browse repositories, commits, branches, pull requests, diffs, and code search
- Builds: List builds, view logs, analyze errors
- Work Items: Query work items, find linked commits/PRs
- Releases: View releases and deployment status
- Pipelines: List YAML pipelines, view runs, get pipeline configuration
- Test Results: View test runs, results, and analyze failures
Prerequisites
- Node.js 18 or higher
- Azure DevOps Server 2020 or 2022 (on-premises)
- Personal Access Token (PAT) with read scopes
Installation
Option 1: npm (Recommended)
npx azure-devops-mcp@latest
Option 2: Clone and Build
git clone https://github.com/elad-nofy/azure-devops-mcp.git
cd azure-devops-mcp
npm install
npm run build
Configuration
Create a Personal Access Token (PAT)
- Go to Azure DevOps Server
- Click on your profile icon > Security > Personal access tokens
- Create a new token with the following scopes:
- Code: Read
- Build: Read
- Work Items: Read
- Release: Read
- Test Management: Read
Configure MCP Client
Claude Code CLI
Add to your global settings (~/.claude.json on macOS/Linux, %USERPROFILE%\.claude.json on Windows):
Windows:
{
"mcpServers": {
"azure-devops": {
"command": "cmd",
"args": ["/c", "npx", "azure-devops-mcp"],
"env": {
"AZURE_DEVOPS_URL": "http://your-tfs-server:8080/tfs",
"AZURE_DEVOPS_PAT": "your-personal-access-token",
"AZURE_DEVOPS_COLLECTION": "YourCollection",
"AZURE_DEVOPS_PROJECT": "YourDefaultProject"
}
}
}
}
macOS / Linux:
{
"mcpServers": {
"azure-devops": {
"command": "npx",
"args": ["azure-devops-mcp"],
"env": {
"AZURE_DEVOPS_URL": "http://your-tfs-server:8080/tfs",
"AZURE_DEVOPS_PAT": "your-personal-access-token",
"AZURE_DEVOPS_COLLECTION": "YourCollection",
"AZURE_DEVOPS_PROJECT": "YourDefaultProject"
}
}
}
}
VS Code (Continue/Cline)
Add similar configuration to the extension's MCP settings.
Environment Variables
| Variable | Required | Description |
|---|---|---|
AZURE_DEVOPS_URL |
Yes | Server base URL (e.g., http://tfs:8080/tfs) |
AZURE_DEVOPS_PAT |
Yes | Personal Access Token |
AZURE_DEVOPS_COLLECTION |
No | Collection name (default: DefaultCollection) |
AZURE_DEVOPS_PROJECT |
No | Default project for commands |
Available Tools
Projects
| Tool | Description |
|---|---|
test_connection |
Test connection to Azure DevOps and verify authentication |
list_projects |
List all projects in the organization |
get_project |
Get detailed project information |
Git / Repositories
| Tool | Description |
|---|---|
list_repos |
List repositories in a project |
list_branches |
List branches in a repository |
list_commits |
Get commit history with filters |
get_commit |
Get commit details with changes |
get_commit_diff |
Get file diff for a commit |
list_pull_requests |
List pull requests |
get_pull_request |
Get PR details with comments |
compare_branches |
Compare two branches (regression analysis) |
search_code |
Search for text in repository files |
get_file_content |
Get file content from a repo at a specific branch/commit |
get_commits_for_work_item |
Find commits referencing a work item |
get_prs_for_work_item |
Find PRs linked to a work item |
Builds
| Tool | Description |
|---|---|
list_builds |
List recent builds |
get_build |
Get build details |
get_build_logs |
Get build logs |
list_build_definitions |
List build definitions |
analyze_build_errors |
Extract errors from failed builds |
Work Items
| Tool | Description |
|---|---|
query_work_items |
Search work items (WIQL or filters) |
get_work_item |
Get work item details |
list_work_item_types |
List available work item types |
list_iterations |
List iterations/sprints in a project |
list_areas |
List area paths in a project |
Releases
| Tool | Description |
|---|---|
list_releases |
List releases |
get_release |
Get release details |
list_release_definitions |
List release definitions |
get_release_logs |
Get deployment logs |
Pipelines
| Tool | Description |
|---|---|
list_pipelines |
List YAML pipelines |
get_pipeline_runs |
Get recent pipeline runs |
get_pipeline_yaml |
Get pipeline YAML configuration |
get_pipeline_variables |
Get pipeline variables |
Test Results
| Tool | Description |
|---|---|
list_test_runs |
List test runs in a project |
get_test_run |
Get test run details |
get_test_results |
Get test results from a run |
get_failed_tests |
Get failed tests from a run |
get_test_runs_for_build |
Get test runs for a specific build |
analyze_test_failures |
Analyze and group test failures |
Usage Examples
Once configured, you can ask your AI assistant questions like:
- "Show me the recent builds for project X"
- "What's the status of PR #123?"
- "List all bugs assigned to me"
- "Show me the commits from last week"
- "Why did build #456 fail?"
- "What tests failed in build #789?"
- "Compare branches Dev/9.1 and Dev/9.2"
- "Search for 'IMediator' in the Controllers folder"
Development
# Install dependencies
npm install
# Build
npm run build
# Watch mode
npm run dev
Troubleshooting
Connection errors
- Verify
AZURE_DEVOPS_URLis correct and accessible - Check if your PAT has not expired
- Ensure your network allows connection to the server
Permission errors
- Verify PAT has required scopes
- Check if you have access to the project/repository
"Project is required" errors
- Set
AZURE_DEVOPS_PROJECTenvironment variable, or - Pass
projectparameter in tool calls
License
MIT
推荐服务器
Baidu Map
百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。
Playwright MCP Server
一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。
Audiense Insights MCP Server
通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。
Magic Component Platform (MCP)
一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。
VeyraX
一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。
Kagi MCP Server
一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。
graphlit-mcp-server
模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。
Exa MCP Server
模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。
mcp-server-qdrant
这个仓库展示了如何为向量搜索引擎 Qdrant 创建一个 MCP (Managed Control Plane) 服务器的示例。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。