Azure DevOps MCP Server

Azure DevOps MCP Server

Focused local MCP server for Azure DevOps Services, Azure Repos, and Azure Boards. It provides ticket-reading and pull-request-review workflows through 31 MCP tools backed by Azure DevOps REST API 7.1.

Category
Visit Server

README

Azure DevOps MCP Server

Local Model Context Protocol (MCP) server for Azure DevOps Services. It exposes 64 tools for Azure Boards, Azure Repos, pull request review and lifecycle management, Azure Pipelines, team sprints/backlogs, and delivery reporting.

Read tools are enabled by default. Every mutation requires both a server-side feature flag and an explicit per-call confirmation. Pull request completion and auto-complete never bypass branch policies.

Capabilities

Azure Boards

  • Read work items, every field, relations, comments, and historical snapshots.
  • Run read-only WIQL SELECT queries and resolve the matching work items.
  • Create and update work items, including state, assignee, tags, area, sprint, and custom fields.
  • Add comments, parent/child/related/dependency/duplicate links, and attachments up to 10 MiB.
  • Use an expected revision to prevent stale work item updates.
  • Trace a ticket to linked pull requests, commits, builds, branches, and related work items.
  • Read the work item's field and relation update history.

Repositories and branches

  • List projects, repositories, clone links, branches, and commits.
  • Create a branch at an exact commit.
  • Delete a branch only when its current object ID matches the caller's expectation.
  • Compare branches with ahead/behind counts and changed items.
  • Find stale branches while excluding protected branch names.

Pull requests and code review

  • List and retrieve pull requests, including direct Azure DevOps PR URLs.
  • Read commits, reviewers/votes, iterations, threads, comments, labels, and linked work items.
  • Generate bounded unified diffs locally and validate exact inline-comment targets.
  • Create and update PRs; switch draft state; abandon or reactivate a PR.
  • Add/remove reviewers and labels; create/reply/resolve review comments; cast votes.
  • Enable policy-respecting auto-complete or merge at an exact reviewed source SHA.
  • Evaluate merge readiness from draft/status, merge state, votes, required reviewers, unresolved threads, policy evaluations, and PR status checks.
  • Produce batch review summaries and stale PR reports.

Azure Pipelines

  • List pipelines, runs, and builds; inspect run state/result.
  • Filter builds by definition, repository, branch, status, and result.
  • Queue or preview a run with a branch, variables, template parameters, and skipped stages.
  • Re-run a full pipeline using a previous run's resolved resources.
  • List build log records and retrieve bounded log text.

Azure DevOps REST does not expose a generic failed-job-only rerun through this implementation; rerun_pipeline queues a full run.

Sprints and backlogs

  • List team iterations by current, past, or future timeframe.
  • Get the work items and team-member capacity for a sprint.
  • Calculate item and story-point velocity with configurable point field and completed states.
  • Reorder or reparent backlog items.
  • Move a work item to a sprint through update_work_item and System.IterationPath.

Tool groups

Group Count Coverage
Core and repository reads 6 Auth, projects, repositories, clone links, branch listing
Azure Boards 8 Work items, WIQL, comments, create/update, relations, attachments
Pull requests and history 10 PR metadata, commits, threads, iterations, reviewers, linked work items
Review and diff 6 Changed files, stats, unified diffs, inline validation, review context
PR review and lifecycle writes 13 PR create/update/merge, auto-complete, comments, votes, reviewers, labels
Branch lifecycle 4 Create, delete, compare, stale branches
Pipelines and builds 7 Definitions, runs, queue/rerun, logs, build filters
Sprints and backlog 5 Iterations, work items, capacity, velocity, reorder
Quality and trace reports 5 Merge readiness, batch/stale reports, delivery trace, audit history
Total 64

The complete input reference is in docs/tools.md.

Requirements

  • Node.js 20+
  • npm
  • An Azure DevOps Services organization
  • An Azure DevOps PAT or an Azure CLI session created with az login

Azure DevOps Server/on-premises is not currently supported.

Quick start

git clone https://github.com/hasanozcan/azure-devops-mcp.git
cd azure-devops-mcp
npm install
Copy-Item .env.example .env

Edit .env:

AZURE_DEVOPS_ORGANIZATION=your-organization
AZURE_DEVOPS_DEFAULT_PROJECT=your-project
AZURE_DEVOPS_AUTH_MODE=pat
AZURE_DEVOPS_PAT=your-token
AZURE_DEVOPS_ENABLE_WRITE_TOOLS=false

Then validate and build:

npm run doctor
npm run build
npm start

The transport is local stdio; stdout is reserved for MCP protocol messages.

Finding organization and project values

For a URL such as:

https://thecellsolutions.visualstudio.com/hpowere/_git/hpower/pullrequests

use:

AZURE_DEVOPS_ORGANIZATION=thecellsolutions
AZURE_DEVOPS_DEFAULT_PROJECT=hpowere

For https://dev.azure.com/<organization>/<project>/..., use the first path segment as the organization and the second as the project.

Authentication and permissions

Recommended PAT permissions depend on the tools you enable:

Capability PAT permission
Repository, branch, commit, PR, and diff reads Code: Read
PR/branch/reviewer/label/comment mutations Code: Read & write
Work item, WIQL, comment, history, sprint-item reads Work Items: Read
Work item creation/update/comments/relations/attachments/backlog reorder Work Items: Read & write
Pipeline, run, build, and log reads Build: Read
Queue, preview, or rerun pipelines Build: Read & execute
Iteration and capacity discovery Project and Team: Read

Use the shortest practical PAT expiration. See docs/auth.md.

Azure CLI mode:

az login
AZURE_DEVOPS_AUTH_MODE=azcli
AZURE_DEVOPS_PAT=

Configuration

Variable Required Default Purpose
AZURE_DEVOPS_ORGANIZATION Yes None Organization name or Azure DevOps URL
AZURE_DEVOPS_DEFAULT_PROJECT No None Project used when a tool omits project
AZURE_DEVOPS_AUTH_MODE No pat pat or azcli
AZURE_DEVOPS_PAT PAT mode None Personal Access Token
AZURE_DEVOPS_BASE_URL No https://dev.azure.com REST API base URL
AZURE_DEVOPS_API_VERSION No 7.1 Default REST API version
AZURE_DEVOPS_ENABLE_WRITE_TOOLS No false Enable guarded mutations
AZURE_DEVOPS_REQUEST_TIMEOUT_MS No 30000 Per-request timeout
AZURE_DEVOPS_RETRY_COUNT No 2 Retry count for safe reads
AZURE_DEVOPS_MAX_DIFF_FILE_BYTES No 1048576 Per-side local diff file limit
AZURE_DEVOPS_MAX_DIFF_LINES No 5000 Maximum returned diff lines

Placeholder secrets such as replace-me are rejected at startup.

Codex configuration

The repository includes .codex/config.toml. Build the project, create .env, trust/open the repository in Codex, and restart the MCP server list.

Manual registration:

codex mcp add azure-devops-mcp -- node "<absolute-path-to-project>\dist\server.js"

The project configuration uses default_tools_approval_mode = "writes", adding Codex approval on top of the server's own mutation gates. .vscode/mcp.json is also included.

Safety model

Every mutation requires:

  1. AZURE_DEVOPS_ENABLE_WRITE_TOOLS=true at startup.
  2. confirm: true in the individual MCP call.

Additional safeguards:

  • work item updates can test expectedRevision;
  • branch deletion requires expectedObjectId;
  • inline comments validate the current iteration, file side, line range, and changeTrackingId;
  • PR completion requires the exact current source commit SHA;
  • PR completion and auto-complete always use bypassPolicy: false;
  • mutation requests are never automatically retried;
  • PATs, bearer tokens, and authorization headers never appear in tool or doctor output.

Example prompts

Read work item 544, including comments, relations, and delivery links.
Move work item 544 to Project\Sprint 8 after I confirm.
Create a branch feature/544 at this exact commit after I confirm.
Review this Azure DevOps pull request URL and report merge blockers.
Add a required reviewer and enable squash auto-complete after I confirm.
Show failed builds for the hpower repository and retrieve the last run logs.
Run pipeline 12 on feature/544 with deploy=false after I confirm.
Show the current sprint capacity and calculated story-point velocity.
Report branches and active PRs older than 45 days.

Validation

npm run typecheck
npm run typecheck:tests
npm test
npm run build
npm run smoke:mcp
npm run doctor
npm pack --dry-run

doctor performs live read-only checks without printing credentials.

Scope boundaries

This release does not manage wikis, test plans, service connections, variable groups, deployment environments, saved queries, repository creation, or Azure DevOps Server/on-premises. Pipeline rerun is full-run only.

API references

The adapters target Microsoft's Azure DevOps REST API 7.1 documentation for Git, Work Item Tracking, Pipelines, Build, and Work/Sprints.

License

MIT © 2026 Hasan Özcan. See LICENSE.

Recommended Servers

playwright-mcp

playwright-mcp

A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.

Official
Featured
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.

Official
Featured
Local
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.

Official
Featured
Local
TypeScript
VeyraX MCP

VeyraX MCP

Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.

Official
Featured
Local
graphlit-mcp-server

graphlit-mcp-server

The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.

Official
Featured
TypeScript
Kagi MCP Server

Kagi MCP Server

An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.

Official
Featured
Python
Neon Database

Neon Database

MCP server for interacting with Neon Management API and databases

Official
Featured
E2B

E2B

Using MCP to run code via e2b.

Official
Featured
Exa Search

Exa Search

A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.

Official
Featured
Qdrant Server

Qdrant Server

This repository is an example of how to create a MCP server for Qdrant, a vector search engine.

Official
Featured