Azure DevOps MCP
A dynamic Azure DevOps MCP (Model Context Protocol) server that automatically switches authentication context based on the current working directory. This enables seamless integration with multiple Azure DevOps organizations and projects from a single MCP server.
Features
- Local Configuration Files: Each repository contains
.azure-devops.json
configuration - Dynamic Environment Switching: Automatically detects project context based on directory location
- Multiple Project Support: Supports unlimited projects with separate authentication
- Comprehensive Azure DevOps Integration: Work items, repositories, builds, and more
- Zero Configuration Switching: Seamless switching between projects with local config files
- Secure Token Storage: PAT tokens stored locally per repository (excluded from git)
- Error Handling & Fallback: Robust error handling with graceful degradation to environment variables
Local Configuration
Each repository should contain a .azure-devops.json
configuration file:
Configuration File Structure
Security Configuration
Important: Add .azure-devops.json
to your .gitignore
file:
Example Projects
RiverSync Project
- Directory:
/Users/wangkanai/Sources/riversync
- Configuration:
.azure-devops.json
with RiverSync organization settings
Mula Project
- Directory:
/Users/wangkanai/Sources/mula
- Configuration:
.azure-devops.json
with Mula organization settings
Installation
Claude Code Installation (Recommended)
Note: The
-y
flag automatically accepts the package installation prompt, ensuring smooth non-interactive execution for MCP servers.
Claude Desktop Installation
For Claude Desktop users, add this configuration to your MCP settings:
Claude Desktop MCP Settings Location:
- macOS:
~/Library/Application Support/Claude/claude_desktop_config.json
- Windows:
%APPDATA%\Claude\claude_desktop_config.json
Alternative Installation Methods
Global Installation
Development Installation
Development
Available Tools
📋 For comprehensive command documentation with detailed examples, see MCP-COMMANDS.md
Work Items
- get-work-items: Retrieve work items using WIQL queries or specific IDs with field selection
- create-work-item: Create new work items with full hierarchy support (Epic → Feature → User Story → Task)
- update-work-item: Update existing work items including state, assignments, parent relationships, and iteration paths
- add-work-item-comment: Add comments to existing work items for progress tracking
Repositories & Pull Requests
- get-repositories: List all repositories in the current project context
- get-pull-requests: Get pull requests with filtering options (status, creator, repository)
Builds & Pipelines
- get-builds: Get build definitions and recent build history with filtering
- trigger-pipeline: Trigger build pipelines with parameters and branch selection
- get-pipeline-status: Get detailed build status and timeline information
Context Management
- get-current-context: Get current Azure DevOps context based on directory
🎯 Key Features
- ✅ Hierarchical Work Items: Full Epic → Feature → User Story → Task hierarchy support
- ✅ Parent Relationships: Establish parent-child relationships during work item creation
- ✅ WIQL Queries: Powerful Work Item Query Language support for complex searches
- ✅ Sprint Management: Iteration path assignment and management
- ✅ Pipeline Integration: Trigger builds and monitor deployment status
- ✅ Multi-Project Support: Seamless switching between Azure DevOps organizations
Usage Examples
Get Current Context
Query Work Items
Create Work Item
Get Repositories
Directory Detection Logic
The server uses intelligent directory detection:
- Exact Match: Direct match of configured directory paths
- Nested Directory Support: Detects parent project directories
- Longest Match Priority: Most specific directory match takes precedence
- Parent Directory Search: Searches up the directory tree for matches
- Fallback Configuration: Uses default configuration when no match found
Local Configuration
Each repository should contain a .azure-devops.json
configuration file:
Configuration File Structure
Security Configuration
Important: Add .azure-devops.json
to your .gitignore
file:
Example Projects
RiverSync Project
- Directory:
/Users/wangkanai/Sources/riversync
- Configuration:
.azure-devops.json
with RiverSync organization settings
Mula Project
- Directory:
/Users/wangkanai/Sources/mula
- Configuration:
.azure-devops.json
with Mula organization settings
Authentication
The server uses Personal Access Tokens (PAT) for Azure DevOps authentication. PAT tokens are configured per project in local .azure-devops.json
configuration files within each repository.
PAT Token Requirements
PAT tokens should have the following scopes:
- Work Items: Read & Write
- Code: Read
- Build: Read
- Project and Team: Read
Error Handling
The server includes comprehensive error handling:
- Configuration Errors: Graceful fallback when configuration is missing
- Authentication Errors: Clear error messages for authentication failures
- API Errors: Detailed error reporting for Azure DevOps API issues
- Network Errors: Retry logic and timeout handling
Testing & Validation
Enhanced Validation System (Recommended)
The enhanced validation system includes MCP server startup, connection verification, and readiness checks:
1. MCP Server Warmup
2. Enhanced Comprehensive Validation
3. Manual Testing
Configuration File
The generic validation system uses validation-config.json
:
Enhanced Validation Features
The enhanced validation system includes:
🚀 MCP Server Management
- Startup Verification: Ensures MCP servers are properly configured
- Connection Testing: Verifies server connectivity with retry logic
- Readiness Checks: Confirms servers respond to basic commands
- Warmup Period: Configurable delay for server initialization (default: 10s)
🔍 Comprehensive Testing Coverage
- ✅ Prerequisites: PowerShell, Claude Code, directory structure, proxy build
- ✅ MCP Initialization: Server startup, connectivity, and readiness verification
- ✅ Local Configuration:
.azure-devops.json
file validation with expected values - ✅ Server Configuration: Local scope verification without environment variables
- ✅ Claude Integration: MCP command execution and context detection
- ✅ Dynamic Switching: Environment switching between multiple repositories
- ✅ Error Handling: Comprehensive error detection and reporting with retry logic
Expected Results
Full Validation Results:
- Pass Rate: >90% for successful implementation
- All MCP commands functional in both repositories
- Automatic context switching based on directory location
Legacy Environment Configuration (Deprecated)
The server previously supported a global config/environments.json
file for environment mapping. This approach has been deprecated in favor of local .azure-devops.json
configuration files for better security and project isolation.
If you need to migrate from the old environment-based configuration, convert your settings to local configuration files in each repository.
Architecture
Core Components
- AzureDevOpsMCPProxy: Main server class with MCP protocol handling
- DirectoryDetector: Smart directory detection and configuration mapping
- ToolHandlers: Azure DevOps API integration and tool implementations
- ConfigLoader: Configuration file loading and validation
Request Flow
- Tool Call Received: MCP client sends tool call request
- Context Detection: Directory detector identifies current project context
- Configuration Switch: Server switches to appropriate Azure DevOps configuration
- API Request: Tool handler makes authenticated API request to Azure DevOps
- Response Processing: Response is formatted and returned to MCP client
Integration with Claude Code
This MCP server is designed to work seamlessly with Claude Code for Azure DevOps operations:
- Automatic Context Switching: When working in RiverSync or Mula project directories
- Transparent Authentication: No manual configuration required
- Rich Tool Set: Comprehensive Azure DevOps functionality
- Error Recovery: Graceful handling of authentication and network issues
Security Considerations
- PAT tokens are stored in configuration files (ensure proper file permissions)
- HTTPS is used for all Azure DevOps API communications
- Authentication is handled per-request with proper token encoding
- No token caching or persistence beyond configuration file
Troubleshooting
Common Issues
1. Installation Command Issues (Issue #14 Resolution)
Problem: Incorrect installation commands that fail to start the server
Root Cause: Outdated documentation showing incorrect command syntax
Solution: Use the correct installation command:
Alternative working commands:
2. Configuration Issues
- Configuration Not Found: Ensure
.azure-devops.json
exists in your project directory - Authentication Errors: Verify PAT token permissions and expiration in your local configuration
- Directory Detection: Check that your project has a valid
.azure-devops.json
file - API Errors: Verify Azure DevOps organization and project names in your local configuration
3. Installation Verification
Test your installation with these commands:
Debug Mode
Enable debug logging by setting environment variable:
NPM Package Technical Details
- Package Name:
@wangkanai/devops-mcp
- Binary Name:
devops-mcp
(auto-generated by NPM) - Latest Version: Check with
npm view @wangkanai/devops-mcp version
- Installation Verification:
npm list -g @wangkanai/devops-mcp
License
MIT License - see LICENSE file for details.
local-only server
The server can only run on the client's local machine because it depends on local resources.
Tools
A dynamic Azure DevOps MCP server that automatically switches authentication context based on the current working directory, enabling seamless integration with multiple Azure DevOps organizations and projects.
- Features
- Local Configuration
- Installation
- Development
- Available Tools
- Usage Examples
- Directory Detection Logic
- Local Configuration
- Authentication
- Error Handling
- Testing & Validation
- Legacy Environment Configuration (Deprecated)
- Architecture
- Integration with Claude Code
- Security Considerations
- Troubleshooting
- License
Related MCP Servers
- AsecurityAlicenseAqualityThis server provides a convenient API for interacting with Azure DevOps services, enabling AI assistants and other tools to manage work items, code repositories, boards, sprints, and more. Built with the Model Context Protocol, it provides a standardized interface for communicating with Azure DevOpsLast updated -963343TypeScriptMIT License
- AsecurityAlicenseAqualityA Model Context Protocol server that enables AI assistants to interact with Azure DevOps resources including projects, work items, repositories, pull requests, branches, and pipelines through a standardized protocol.Last updated -15506285TypeScriptMIT License
- AsecurityAlicenseAqualityA Model Context Protocol server that enables AI assistants to interact with Azure DevOps services, allowing users to query work items with plans to support creating/updating items, managing pipelines, handling pull requests, and administering sprints and branch policies.Last updated -972PythonMIT License
- -securityAlicense-qualityAn MCP server for interacting with Azure. Contains some common Compute and Networking actions, and extensible to add many more.Last updated -1PythonApache 2.0