Azure DevOps MCP Server
Click on "Deploy Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type
@followed by the MCP server name and your instructions, e.g., "@Azure DevOps MCP Server Show me my open pull requests"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Azure DevOps MCP Server
An MCP (Model Context Protocol) server that exposes 41 Azure DevOps tools to AI assistants — covering pipelines, Git repositories, pull requests, releases, boards, work items, test management, and wikis.
Authentication: Azure Entra device code flow — no app registration required.
What It Can Do
Category | Tools | Highlights |
Pipelines & Builds | 6 | List pipelines, view runs/logs, trigger builds, browse build artifacts |
Releases & Approvals | 6 | List/create releases, view release details, list/approve/reject approvals |
Git Repositories | 6 | List repos, browse branches, view commits, read file content, compare branches |
Pull Requests | 6 | List/create/update PRs, read comment threads, add comments (inline or general) |
Boards & Work Items | 10 | List boards, get/create/update/delete work items, comments, WIQL queries, saved queries |
Test Management | 3 | List test runs, view test results, get code coverage |
Wikis | 3 | List wikis, read pages, create/update pages |
41 total |
Related MCP server: ado-mcp
Prerequisites
Python 3.10+
Access to an Azure DevOps organization (dev.azure.com)
Installation
From PyPI (recommended)
pip install azure-devops-mcp-serverOr with uv:
uv pip install azure-devops-mcp-serverFrom source
git clone <your-repo-url>
cd azure-devops-mcp
pip install -e .Quick Start
After installing, follow these 3 steps:
Step 1: Install
pip install azure-devops-mcp-serverStep 2: Add to your AI editor
Open your editor's MCP config file and add the server. Here's VS Code as an example (.vscode/mcp.json):
Option A — With a PAT (simplest, no browser prompt):
{
"servers": {
"azure-devops": {
"command": "azure-devops-mcp",
"env": {
"AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/YOUR-ORG",
"AZURE_DEVOPS_PAT": "YOUR-PERSONAL-ACCESS-TOKEN"
}
}
}
}To create a PAT: Azure DevOps → User Settings → Personal Access Tokens → New Token.
Option B — With Device Code Flow (no PAT needed):
First, log in once from your terminal:
export AZURE_DEVOPS_ORG_URL=https://dev.azure.com/YOUR-ORG
azure-devops-mcp auth loginThis opens your browser, you sign in, and the token is cached locally. Then add to your editor — no PAT required:
{
"servers": {
"azure-devops": {
"command": "azure-devops-mcp",
"env": {
"AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/YOUR-ORG"
}
}
}
}Token is cached at
~/.azure-devops-mcp/token_cache.jsonand auto-refreshes (~90 days). Runazure-devops-mcp auth statusto check your login state.
Step 3: Use it
Open your AI assistant (Copilot Chat, Claude, etc.) and ask:
List all pipelines in my projectThe server handles authentication and API calls — the AI tool sees 41 Azure DevOps commands automatically.
Configuration
Set these environment variables (via your shell, .env file, or MCP client config):
Variable | Required | Description |
| Yes | Your Azure DevOps org URL (e.g., |
| No | Default project (can be overridden per tool call) |
Authentication
The server supports 4 authentication methods, auto-detected from environment variables:
Method 1: Personal Access Token (PAT) — simplest
Best for: scripts, personal use, CI/CD pipelines.
Variable | Description |
|
"env": {
"AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/my-org",
"AZURE_DEVOPS_PAT": "your-pat-here"
}Method 2: Service Principal (Client Credentials) — for apps
Best for: registered Azure AD applications, service-to-service, automated workflows.
Variable | Description |
| Application (client) ID from your Azure AD app registration |
| Client secret for the app |
| Azure AD tenant ID |
"env": {
"AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/my-org",
"AZURE_CLIENT_ID": "your-app-id",
"AZURE_CLIENT_SECRET": "your-secret",
"AZURE_TENANT_ID": "your-tenant-id"
}Method 3: Managed Identity — for Azure-hosted environments
Best for: running on Azure VMs, App Service, Azure Functions, AKS. No credentials to manage.
pip install azure-devops-mcp-server[managed-identity]Variable | Description |
| Set to |
| (optional) Client ID for user-assigned managed identity |
"env": {
"AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/my-org",
"AZURE_USE_MANAGED_IDENTITY": "true"
}Method 4: Device Code Flow — interactive (default)
Best for: developers running locally. No PAT or app registration needed.
Recommended: Pre-authenticate from your terminal before configuring your AI editor:
export AZURE_DEVOPS_ORG_URL=https://dev.azure.com/my-org
azure-devops-mcp auth login # Opens browser, caches token
azure-devops-mcp auth status # Verify login stateThe browser opens automatically. After you sign in, the token is cached at ~/.azure-devops-mcp/token_cache.json and auto-refreshes (~90 days).
If you skip auth login, the server will prompt via stderr on the first API call — but most AI editors don't display stderr, so pre-authenticating is strongly recommended.
Priority: PAT → Client Credentials → Managed Identity → Device Code. The first method detected wins.
Setting Environment Variables
There are 3 ways to pass your configuration:
Option A — In your AI editor's MCP config (recommended)
Every MCP-compatible editor has an "env" block in its config. This is the easiest approach — just add your variables there. See the Integration section below for editor-specific examples.
Option B — Via a .env file
Create a .env file in your working directory (use .env.example as a template):
AZURE_DEVOPS_ORG_URL=https://dev.azure.com/my-org
AZURE_DEVOPS_PAT=my-pat-tokenThe server loads it automatically on startup via python-dotenv.
Option C — Via shell environment (CI/CD, Docker, manual runs)
export AZURE_DEVOPS_ORG_URL=https://dev.azure.com/my-org
export AZURE_DEVOPS_PAT=my-pat-token
azure-devops-mcpIntegration
After installing via pip, the azure-devops-mcp command is available on your PATH. Configure your AI tool:
VS Code (GitHub Copilot / Copilot Chat)
Add to .vscode/mcp.json in your workspace (or User Settings JSON under mcp.servers):
{
"servers": {
"azure-devops": {
"command": "azure-devops-mcp",
"env": {
"AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/your-organization",
"AZURE_DEVOPS_PROJECT": "your-default-project"
}
}
}
}Claude Desktop
Add to your Claude Desktop config file:
Windows:
%APPDATA%\Claude\claude_desktop_config.jsonmacOS:
~/Library/Application Support/Claude/claude_desktop_config.json
{
"mcpServers": {
"azure-devops": {
"command": "azure-devops-mcp",
"env": {
"AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/your-organization",
"AZURE_DEVOPS_PROJECT": "your-default-project"
}
}
}
}Claude Code (Cortex Code)
claude mcp add azure-devops -- azure-devops-mcpThen set environment variables in your shell or Claude config.
Cursor
Add to .cursor/mcp.json in your workspace:
{
"mcpServers": {
"azure-devops": {
"command": "azure-devops-mcp",
"env": {
"AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/your-organization",
"AZURE_DEVOPS_PROJECT": "your-default-project",
"AZURE_TENANT_ID": "your-tenant-id-or-organizations"
}
}
}
}Windsurf
Add to ~/.codeium/windsurf/mcp_config.json:
{
"mcpServers": {
"azure-devops": {
"command": "azure-devops-mcp",
"env": {
"AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/your-organization",
"AZURE_DEVOPS_PROJECT": "your-default-project"
}
}
}
}Any MCP-compatible client
The server communicates over stdio. Point any MCP client at:
azure-devops-mcpRunning from source (alternative)
If you cloned the repo instead of installing via pip:
{
"servers": {
"azure-devops": {
"command": "uv",
"args": ["run", "python", "-m", "azure_devops_mcp"],
"cwd": "/path/to/azure-devops-mcp",
"env": {
"AZURE_DEVOPS_ORG_URL": "https://dev.azure.com/your-organization",
"AZURE_DEVOPS_PROJECT": "your-default-project"
}
}
}
}All 41 Tools
Pipelines & Build Artifacts
Tool | Description |
| List all YAML pipelines in a project |
| Get recent runs for a pipeline |
| Fetch full logs for a pipeline run |
| Trigger a new pipeline run with optional template parameters |
| List artifacts produced by a build |
| Get the download URL for a specific build artifact |
Releases & Approvals
Tool | Description |
| List classic release definitions (release pipelines) |
| List releases, optionally filtered by definition |
| Get release detail including environment/stage status |
| Create a release with optional artifact version overrides |
| List pending (or other status) release approvals |
| Approve or reject a release approval |
Git Repositories
Tool | Description |
| List all Git repositories in a project |
| Get detailed info about a repository (URLs, default branch, etc.) |
| List branches, optionally filtered by prefix |
| Get recent commits, optionally filtered by branch or author |
| Read a file's content from a repository |
| Compare two branches (ahead/behind count + changed files) |
Pull Requests
Tool | Description |
| List PRs by status (active/completed/abandoned) |
| Get PR detail (reviewers, merge status, etc.) |
| Create a new PR with optional reviewers and draft mode |
| Update PR status, title, description, auto-complete, merge strategy |
| Get all comment threads on a PR |
| Add a comment (general or inline on a specific file/line) |
Boards & Work Items
Tool | Description |
| List boards for a project/team |
| Get work items on a board with column info |
| Get full work item detail (fields, relations, estimates) |
| Create a Bug, User Story, Task, Epic, or Feature |
| Update fields and/or add relation links |
| Delete a work item (recycle bin or permanent) |
| Execute an arbitrary WIQL query |
| Get comments on a work item |
| Add a comment to a work item |
| List saved query folders and queries |
| Execute a saved query by its ID |
Test Management
Tool | Description |
| List test runs, optionally filtered by state |
| Get test results for a run (pass/fail/error details) |
| Get code coverage summary for a build |
Wikis
Tool | Description |
| List all wikis in a project |
| Get a wiki page's content by path |
| Create or update a wiki page with markdown content |
Example Prompts
List all pipelines in the "MyProject" projectShow me the last 5 runs for pipeline 42List all active pull requests in the "backend-api" repositoryCreate a pull request from feature/login to main titled "Add login page"Show me pending release approvalsApprove release approval 1234 with comment "Looks good, deploying to prod"Query all active bugs assigned to me:
SELECT [System.Id], [System.Title] FROM workitems
WHERE [System.WorkItemType] = 'Bug' AND [System.State] = 'Active'
AND [System.AssignedTo] = @MeGet the wiki page at /Runbooks/Deployment from the project wikiShow test results for test run 567 — only failuresCompare the main and develop branches in the "frontend" repoDevelopment
Run the MCP dev inspector to test tools interactively:
uv run mcp dev src/azure_devops_mcp/server.pyArchitecture
azure-devops-mcp/
├── src/azure_devops_mcp/
│ ├── __init__.py # Package init + version
│ ├── __main__.py # CLI entry point (azure-devops-mcp command)
│ ├── server.py # MCP server — registers all 41 tools
│ ├── auth.py # MSAL device-code auth (no app registration needed)
│ └── clients/
│ ├── pipelines.py # Pipeline + build artifact API calls
│ ├── releases.py # Classic release + approval API calls
│ ├── git.py # Git repos, branches, commits, PRs
│ ├── boards.py # Boards, work items, comments, saved queries
│ ├── tests.py # Test runs, results, code coverage
│ └── wiki.py # Wiki pages
├── .env.example # Environment variable template
├── pyproject.toml # Python project config (PyPI metadata + build)
└── README.mdAll client modules include retry logic with exponential backoff for 429 Too Many Requests and 5xx server errors.
License
Available Tools
41 toolsadd_work_item_commentA
Add a comment to a work item.
Args: project: Azure DevOps project name. Uses default if not specified. work_item_id: The ID of the work item. text: The comment text (supports HTML).
Inputs: project, work_item_id, text.
compare_branchesB
Compare two branches and show the diff summary.
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name. base_branch: The base branch name. target_branch: The target branch name.
Inputs: project, repository_id, base_branch, target_branch.
create_or_update_wiki_pageA
Create or update a wiki page.
Args: project: Azure DevOps project name. Uses default if not specified. wiki_id: The wiki identifier (name or ID). path: Page path (e.g., "/Release Notes/v3"). content: Markdown content for the page. comment: Optional commit comment. if_match: ETag for optimistic concurrency (required for updates, empty for creates).
Inputs: project, wiki_id, path, content, +2 more.
create_pull_requestA
Create a new pull request.
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name. source_branch: Source branch name (e.g., "feature/my-feature"). target_branch: Target branch name (e.g., "main"). title: Title of the pull request. description: Optional description/body of the pull request. reviewers: Optional list of reviewer IDs. is_draft: Whether to create as a draft PR (default: False).
Inputs: project, repository_id, source_branch, target_branch, +4 more.
create_pull_request_commentA
Add a comment thread to a pull request.
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name. pull_request_id: The pull request ID. content: The comment text (supports markdown). file_path: Optional file path for inline comments. line_number: Optional line number for inline comments (requires file_path). status: Thread status: "active", "fixed", "wontFix", "closed", "byDesign", "pending".
Inputs: project, repository_id, pull_request_id, content, +3 more.
create_releaseA
Create (trigger) a new release from a classic release definition.
Args: project: Azure DevOps project name. Uses default if not specified. definition_id: The release definition ID to create a release from. description: Optional description for the release. artifacts: Optional list of artifact version overrides. Each dict should have: - alias (str): artifact source alias - version_id (str): specific version/build ID - version_name (str): display name for the version
Inputs: project, definition_id, description, artifacts.
create_work_itemA
Create a new work item.
Args: project: Azure DevOps project name. Uses default if not specified. work_item_type: Type of work item (e.g., "Bug", "User Story", "Task", "Epic", "Feature"). title: Title of the work item. description: HTML description of the work item. assigned_to: Display name or email of the person to assign to. area_path: Area path (e.g., "MyProject\Team A"). iteration_path: Iteration path (e.g., "MyProject\Sprint 1"). priority: Priority (1=Critical, 2=High, 3=Medium, 4=Low). tags: Semicolon-separated tags (e.g., "frontend; urgent"). additional_fields: Dict of additional field paths and values. parent_id: Optional ID of a parent work item to link to.
Inputs: project, work_item_type, title, description, +7 more.
delete_work_itemA
Delete a work item (move to recycle bin, or permanently destroy).
Args: project: Azure DevOps project name. Uses default if not specified. work_item_id: The ID of the work item to delete. destroy: If True, permanently delete. If False (default), move to recycle bin.
Inputs: project, work_item_id, destroy.
get_artifact_download_urlB
Get the download URL for a specific build artifact.
Args: project: Azure DevOps project name. Uses default if not specified. build_id: The build ID. artifact_name: The name of the artifact.
Inputs: project, build_id, artifact_name.
get_board_work_itemsC
Get work items on a specific board along with board columns.
Args: project: Azure DevOps project name. Uses default if not specified. board: Name of the board (e.g., "Stories", "Bugs"). team: Team name. If not specified, uses the project's default team.
Inputs: project, board, team.
get_code_coverageB
Get code coverage summary for a build.
Args: project: Azure DevOps project name. Uses default if not specified. build_id: The build ID to get coverage for.
Inputs: project, build_id.
get_commitsA
Get recent commits for a Git repository.
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name. branch: Optional branch name to filter commits. top: Maximum number of commits to return (default: 20). author: Optional author name to filter commits.
Inputs: project, repository_id, branch, top, +1 more. Returns: result
get_file_contentA
Get the content of a file from a Git repository.
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name. path: File path in the repository (e.g., "/src/main.py"). branch: Optional branch name (defaults to the repository's default branch).
Inputs: project, repository_id, path, branch. Returns: result
get_pipeline_run_logsC
Get logs for a specific pipeline run.
Args: project: Azure DevOps project name. Uses default if not specified. pipeline_id: The ID of the pipeline. run_id: The ID of the pipeline run.
Inputs: project, pipeline_id, run_id. Returns: result
get_pipeline_runsA
Get recent runs for a specific pipeline.
Args: project: Azure DevOps project name. Uses default if not specified. pipeline_id: The ID of the pipeline. top: Maximum number of runs to return (default: 10).
Inputs: project, pipeline_id, top. Returns: result
get_pull_requestB
Get detailed information about a specific pull request.
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name. pull_request_id: The pull request ID.
Inputs: project, repository_id, pull_request_id.
get_pull_request_threadsB
Get comment threads on a pull request.
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name. pull_request_id: The pull request ID.
Inputs: project, repository_id, pull_request_id. Returns: result
get_releaseA
Get detailed information about a specific release, including environment/stage status.
Args: project: Azure DevOps project name. Uses default if not specified. release_id: The ID of the release.
Inputs: project, release_id.
get_repositoryA
Get detailed information about a specific Git repository.
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name.
Inputs: project, repository_id.
get_test_run_resultsA
Get test results for a specific test run.
Args: project: Azure DevOps project name. Uses default if not specified. run_id: The test run ID. top: Maximum number of results to return (default: 200). outcome: Optional filter by outcome (e.g., "Passed", "Failed", "NotExecuted").
Inputs: project, run_id, top, outcome. Returns: result
get_wiki_pageA
Get a wiki page by path.
Args: project: Azure DevOps project name. Uses default if not specified. wiki_id: The wiki identifier (name or ID). path: Page path (e.g., "/Home", "/Release Notes/v2"). Defaults to root. include_content: Whether to include the page markdown content (default: True).
Inputs: project, wiki_id, path, include_content.
get_work_itemB
Get detailed information about a specific work item.
Args: project: Azure DevOps project name. Uses default if not specified. work_item_id: The ID of the work item.
Inputs: project, work_item_id.
get_work_item_commentsA
Get comments on a work item.
Args: project: Azure DevOps project name. Uses default if not specified. work_item_id: The ID of the work item. top: Maximum number of comments to return (default: 50).
Inputs: project, work_item_id, top. Returns: result
list_boardsA
List all boards for a project/team.
Args: project: Azure DevOps project name. Uses default if not specified. team: Team name. If not specified, uses the project's default team.
Inputs: project, team. Returns: result
list_branchesB
List branches in a Git repository.
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name. filter_prefix: Optional prefix to filter branch names (e.g., "feature/").
Inputs: project, repository_id, filter_prefix. Returns: result
list_build_artifactsC
List artifacts produced by a build.
Args: project: Azure DevOps project name. Uses default if not specified. build_id: The build ID.
Inputs: project, build_id. Returns: result
list_pipelinesA
List all pipelines in an Azure DevOps project.
Args: project: Azure DevOps project name. Uses default if not specified.
Inputs: project. Returns: result
list_pull_requestsA
List pull requests in a Git repository.
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name. status: Filter by status: "active", "completed", "abandoned", "all" (default: "active"). top: Maximum number of pull requests to return (default: 25). creator: Optional filter by creator ID. reviewer: Optional filter by reviewer ID.
Inputs: project, repository_id, status, top, +2 more. Returns: result
list_release_approvalsA
List release approvals, optionally filtered by status.
Args: project: Azure DevOps project name. Uses default if not specified. status: Filter by status: "pending", "approved", "rejected", "reassigned", "canceled", "skipped", "undefined". Default is "pending". top: Maximum number of approvals to return (default: 25). assigned_to: Optional filter by approver display name or unique name.
Inputs: project, status, top, assigned_to. Returns: result
list_release_definitionsA
List all classic release definitions (release pipelines) in a project.
Args: project: Azure DevOps project name. Uses default if not specified. top: Maximum number of definitions to return (default: 50).
Inputs: project, top. Returns: result
list_releasesA
List releases, optionally filtered by a release definition.
Args: project: Azure DevOps project name. Uses default if not specified. definition_id: Optional release definition ID to filter by. 0 means all. top: Maximum number of releases to return (default: 25).
Inputs: project, definition_id, top. Returns: result
list_repositoriesA
List all Git repositories in a project.
Args: project: Azure DevOps project name. Uses default if not specified.
Inputs: project. Returns: result
list_saved_queriesA
List saved work item queries (folders and queries).
Args: project: Azure DevOps project name. Uses default if not specified. depth: How deep to recurse into query folders (default: 2).
Inputs: project, depth. Returns: result
list_test_runsB
List test runs in a project.
Args: project: Azure DevOps project name. Uses default if not specified. top: Maximum number of runs to return (default: 25). state: Optional filter by state (e.g., "Completed", "InProgress", "Aborted").
Inputs: project, top, state. Returns: result
list_wikisA
List all wikis in a project.
Args: project: Azure DevOps project name. Uses default if not specified.
Inputs: project. Returns: result
query_work_itemsA
Execute a WIQL (Work Item Query Language) query.
Args: project: Azure DevOps project name. Uses default if not specified. query: WIQL query string. Example: "SELECT [System.Id], [System.Title], [System.State] FROM workitems WHERE [System.WorkItemType] = 'Bug' AND [System.State] = 'Active' ORDER BY [System.CreatedDate] DESC"
Inputs: project, query. Returns: result
run_saved_queryB
Execute a saved query by its ID and return work item details.
Args: project: Azure DevOps project name. Uses default if not specified. query_id: The saved query GUID.
Inputs: project, query_id. Returns: result
trigger_pipelineA
Trigger a new pipeline run.
Args: project: Azure DevOps project name. Uses default if not specified. pipeline_id: The ID of the pipeline to trigger. parameters: Optional template parameters to pass to the pipeline as key-value pairs.
Inputs: project, pipeline_id, parameters.
update_pull_requestB
Update a pull request (change status, title, description, set auto-complete).
Args: project: Azure DevOps project name. Uses default if not specified. repository_id: The repository ID or name. pull_request_id: The pull request ID. status: "active", "abandoned", or "completed" (to merge). title: New title for the pull request. description: New description for the pull request. auto_complete_set_by: User ID to set auto-complete by. merge_strategy: "noFastForward", "squash", "rebase", "rebaseMerge".
Inputs: project, repository_id, pull_request_id, status, +4 more.
update_release_approvalB
Approve or reject a release approval.
Args: project: Azure DevOps project name. Uses default if not specified. approval_id: The approval ID. status: "approved" or "rejected". comments: Optional comments for the approval decision.
Inputs: project, approval_id, status, comments.
update_work_itemB
Update an existing work item's fields and/or add relation links.
Args: project: Azure DevOps project name. Uses default if not specified. work_item_id: The ID of the work item to update. fields: Dict mapping field names to new values. Example: {"System.Title": "New Title", "System.State": "Active", "System.AssignedTo": "user@example.com"} add_links: List of relation links to add. Each dict must have: - target_id (int): ID of the work item to link to. - link_type (str): "parent", "child", "related", "predecessor", "successor", or a fully-qualified relation type. - comment (str, optional): Comment for the link. Example: [{"target_id": 100, "link_type": "parent"}]
Inputs: project, work_item_id, fields, add_links.
Tool Schema Changelog
Recent tool additions, removals, and schema changes observed during successful MCP inspections.
Loading tool schema changelog…
TDQS
Scored across 41 tools
Each tool targets a distinct resource-action combination with clear boundaries. For example, work items have separate tools for get, create, update, delete, comments, and query, all with unambiguous names. No two tools appear to do the same thing.
Tool names consistently follow a verb_noun pattern (e.g., create_work_item, list_repositories, get_pull_request). The naming style is uniform across all 41 tools, using snake_case and predictable verbs like 'get', 'list', 'create', 'update', 'delete', etc.
41 tools is on the high side for a single server. While the server covers a broad domain (Azure DevOps), the number of tools exceeds the typical well-scoped range of 3-15. It is borderline but still manageable for an agent, though a modular breakdown might be cleaner.
The tool surface covers the major Azure DevOps areas: work items, git repos, pull requests, pipelines, releases, wikis, boards, and test runs. Minor gaps exist, such as no tool to create a pull request thread directly (only add comments) and no tool to update a release definition. Overall, the core workflows are well-supported.
Maintenance
Related MCP Connectors
The MCP server for Azure DevOps, bringing the power of Azure DevOps directly to your agents.
MCP server for your apps' tools and custom tools, plus hosted AI agents and approval-gated workflows
MCP server that lets your AI assistant build and deploy apps, hosting each one at a live URL.
One MCP server for the services your AI needs: memory, files, Microsoft 365, kanban and more.
Related MCP Servers
- FlicenseAqualityDmaintenanceAn MCP server that provides seamless interaction with Azure DevOps Git repositories, enabling users to manage repositories, branches, pull requests, and pipelines through natural language.83-
- AlicenseNot gradedqualityDmaintenanceAn MCP server that provides Azure DevOps integration for AI assistants, enabling them to list projects, run pipelines, analyze failures, view logs, and troubleshoot builds directly from chat interfaces.MIT
- AlicenseNot gradedqualityCmaintenanceMCP server that enables AI assistants to manage Azure DevOps resources including Work Items, Git repositories, pipelines, and user identities.MIT
- AlicenseNot gradedqualityCmaintenanceEnables AI assistants to interact with Azure DevOps through an MCP server with 34 tools for PRs, work items, repos, and branches, optimized for token efficiency and agent-first workflows.4 npmMIT