GitLab MCP. Manage projects, issues, and pipelines from your chat.

Q: How do I get a Personal Access Token (PAT) for GitLab?

Log in to GitLab, go to User Settings > Access Tokens, and you can generate a new token with the 'api' scope there.

Q: Can I see pipeline failures via the agent?

Yes! Use the 'listprojectpipelines' tool to retrieve the status of all recent pipelines, allowing you to quickly identify failed builds.

Q: How do I use the listvisibleprojects tool to find a project?

The listvisibleprojects tool returns a list of all projects the authenticated user can access. You can then use the project name or ID to feed into other tools, like getprojectdetails, for deeper analysis.

Claude

ChatGPT

Cursor

Gemini

Windsurf

VS Code

JetBrains

Vercel

See Vinkius in Action

Works with every AI agent you already use

…and any MCP-compatible client

Just plug in your AI agents and start using Vinkius.

GitLab MCP Server connects your entire development ecosystem to your AI client. Use it to list projects, check CI/CD pipeline status, track open issues, and read file contents across your entire GitLab instance.

It lets your agent manage the full DevSecOps lifecycle—from initial issue creation to final deployment—all via natural conversation. It's your central hub for project metadata and code visibility.

What your AI agents can do

Create merge request

Requires title.

Create a new merge request

Create project issue

Opens a new issue in a specific GitLab project.

Get merge request

Use the IID (the MR number shown in the UI, not the internal ID).

Get details for a specific merge request

+ 13 more capabilities included

Discover Project Metadata

Retrieve configuration details and metadata for specific projects across your GitLab instance.

Manage Issues and Merge Requests

List, track, and programmatically create project issues and merge requests via chat commands.

Monitor CI/CD Pipelines

Fetch a list of pipelines for a project, allowing you to check build and deployment status in real-time.

Inspect Codebase Files

Read the actual contents of files within any repository, letting you understand code structure or documentation without navigating the UI.

Search Across the Instance

Run global searches that pull results from projects, issues, and users across your entire GitLab setup.

Verify User Identity

Access detailed profile information for the authenticated user to check permissions or account context.

Ask AI about this MCP

Ask ChatGPT

Ask Claude

Ask Perplexity

Supported MCP Clients

OAuth 2.0 Compatible

Claude

ChatGPT

Cursor

Gemini

VS Code

JetBrains

Vercel

Zendesk

+ other MCP clients

Free for Subscribers

Waiting for input…

AI Agent

GitLab MCP Server: 12 Tools for DevSecOps

Use these tools to manage project metadata, track issues, and monitor CI/CD pipelines across your entire GitLab instance via natural conversation.

Make your AI actually useful.

Add this MCP to Claude, Cursor, or Windsurf and your AI stops guessing. It gets real tools to look things up, take action, and handle the stuff you keep doing by hand.

Start using GitLab on Vinkius

create019e9a98

create merge request

Requires title. Create a new merge request

create019d75a5

create project issue

Opens a new issue in a specific GitLab project.

get019e9a98

get merge request

Use the IID (the MR number shown in the UI, not the internal ID). Get details for a specific merge request

get019d75a5

get my gitlab profile

Retrieves the identity and profile details of the currently logged-in user.

get019d75a5

get project details

Fetches the configuration metadata for a specific GitLab project.

get019d75a5

get repository file

Reads the text content of a specified file within a repository.

list019e9a98

list branches

List all branches in a project repository

list019d75a5

list merge requests

Lists all open or merged merge requests for a project.

list019d75a5

list project forks

Lists all forks associated with a specific project.

list019d75a5

list project issues

Lists all open or closed issues for a project.

list019e9a98

list project members

List all members of a project

list019d75a5

list project pipelines

Lists the most recent CI/CD pipelines for a project to check status.

list019d75a5

list visible groups

Lists all GitLab groups the user has access to.

list019d75a5

list visible projects

Lists all GitLab projects the user has access to.

search019d75a5

search gitlab global

Runs a comprehensive search across all projects, issues, and users in GitLab.

verify019d75a5

verify api connection

Checks if the connection between the AI client and GitLab is active and authorized.

Choose How to Get Started

Build a custom MCP for your own tools, or connect a ready-made integration from our catalog.

Build Your Own

Turn any API into an MCP. Import a spec, define Agent Skills, or deploy with MCPFusion.

Import from OpenAPI, Swagger, or YAML specs
Create Agent Skills with progressive disclosure
Deploy to edge with MCPFusion framework
Built in DLP, auth, and compliance on every call
Real time usage dashboard and cost metering
Publish to catalog or keep private

Start building

Make Your AI Do More

Start with GitLab, then connect any of our 4,800+ other servers whenever your AI needs more. One click, no limits.

Use this MCP plus 4,800+ others, all in one place
Add new capabilities to your AI anytime you want
Every connection is secured and compliant automatically
Track usage and costs across all your servers
Works with Claude, ChatGPT, Cursor, and more
New servers added to the catalog every week

Independent Platform Disclaimer: Vinkius is an independent platform and is not affiliated with, endorsed by, sponsored by, verified by, or otherwise authorized by GitLab. All third-party trademarks, logos, and brand names are the property of their respective owners. Their use on this website is strictly for informational purposes to identify service compatibility and interoperability.

VINKIUS INFRASTRUCTURE

Cloud Hosted

Managed infra

V8 Isolated

Sandboxed per request

Zero-Trust Proxy

No stored credentials

DLP Enforced

Policy on every call

GDPR Compliant

EU data residency

Token Compression

~60% cost reduction

Your data is protected. See how we built it.

Works with Claude, ChatGPT, Cursor, and more

The Model Context Protocol standardizes how applications expose capabilities to LLMs. Instead of operating in isolation, your AI gains direct access to external platforms, live data, and real-world actions through secure, standardized connections.

This server provides 16 capabilities that interface natively with Claude, ChatGPT, Cursor, and any MCP client. No middleware. No custom integration required.

Tracking project status shouldn't require 10 clicks.

Today, checking a project's status means opening the main dashboard, navigating to the Issues tab to check blockers, then clicking to the Merge Requests section to see code status. You then have to switch tabs again to the CI/CD section just to see if the latest build passed. It's a mess of context switching and manual data correlation.

With this MCP server, you just ask: 'What's the status of the auth fix MR?' The agent runs `list_merge_requests` and `list_project_pipelines` simultaneously, giving you a single, definitive answer right in your chat window.

GitLab MCP Server: Get the full project picture from your chat.

The biggest time sink goes away: manually cross-referencing a bug report's ID (from the Issues tab) with the correct branch (from the MR tab) and then running a manual pipeline check. You stop hopping between tabs.

Now, you can ask complex questions—like 'Show me all open issues in the 'Infra' group that haven't been linked to a pipeline in the last 24 hours.' The system handles the lookup; you just get the answer.

Support 24/7 support@vinkius.com ↗

Security Vinkius Trust Center ↗

SLA Service Level Agreement ↗

Report Listing Send Report ↗

What you can do with this MCP connector

This server connects your entire development stack to your AI client. It lets your agent manage the full DevSecOps lifecycle—from opening issues to deploying code—all through natural conversation. You can check project metadata, track issues, monitor pipelines, and read file contents across your whole GitLab instance.

To start, your agent first runs verify_api_connection to confirm the link to GitLab is active and authorized. You can check your own profile details by calling get_my_gitlab_profile, which retrieves the identity and profile details of the user logged into GitLab.

You can see what projects and groups are available by running list_visible_groups or list_visible_projects, and you can find out the specific configuration metadata for any project using get_project_details. When you need to search across everything, search_gitlab_global runs a comprehensive search pulling results from projects, issues, and users across your entire GitLab setup.

For managing project tasks, you can list all open or closed issues for a project with list_project_issues, or open a new issue using create_project_issue. You can check all open or merged merge requests for a project with list_merge_requests. You can also list all forks associated with a project by calling list_project_forks.

To monitor deployments, list_project_pipelines lists the most recent CI/CD pipelines for a project, letting you check build and deployment status. You can read the actual contents of any file in a repository using get_repository_file, which reads the text content of a specified file. Finally, you can find everything you need by running search_gitlab_global.

Built · Hosted · Managed by Vinkius GitLab MCP Server - Manage Projects & Pipelines Server ID 019d75a5-7114-73e7-b702-5fec37cea5ba

Vinkius Inspector

Compliance Grade A+

Score 100/100

Report View Report ↗

How GitLab MCP Works

1 Subscribe to the GitLab server via the Vinkius Marketplace and provide your GitLab Base URL and Personal Access Token.
2 Your AI client uses the MCP to call a specific tool, like list_project_pipelines.
3 The server executes the API call against GitLab and returns the structured data (e.g., pipeline status, issue list) directly to your AI client.

The bottom line is, your AI client talks to GitLab, and the server translates that conversation into actionable data.

What Changes When You Connect

Check build health instantly. Instead of clicking into a project dashboard to check the CI/CD status, calling list_project_pipelines gives you the latest pipeline status immediately.
Audit code on demand. Need to see what was in README.md from six months ago? Use get_repository_file to read the content directly, bypassing file tree navigation.
Centralize project knowledge. Use search_gitlab_global to find a specific issue or project across the entire instance, instead of navigating through group hierarchies.
Streamline issue tracking. Instead of manually creating a ticket in the UI, use create_project_issue to open a new issue directly from your chat prompt.
Track development flow. Quickly list all open merge requests with list_merge_requests to understand which code branches are currently under review.
Verify context. Use get_my_gitlab_profile to confirm your user permissions or check account context before running critical commands.

Real-World Use Cases

Finding the Status of a Stalled Feature

A developer needs to know why Feature X hasn't been merged. They ask their agent to check the status. The agent calls list_project_pipelines (to see if the last build failed) and then list_merge_requests (to see if the MR is stuck). The developer gets the full picture without leaving their terminal.

Onboarding a New Team Member

A project manager needs to hand off project context to a new hire. They ask the agent to list all visible projects (list_visible_projects) and then retrieve the main project metadata (get_project_details). The agent compiles a digestible report, saving hours of manual documentation gathering.

Compliance Audit of Codebase

A security engineer needs to verify the version of a specific dependency. They prompt the agent to read a file (get_repository_file) in the core repository. The agent retrieves the file content, allowing the engineer to verify the version number instantly for auditing purposes.

Tracking Down a Specific Bug Report

A tech lead remembers a specific bug report but can't find the issue number. They ask the agent to search globally using search_gitlab_global with keywords like 'auth bug' and the project name. The agent returns the link and details of the matching issue.

The Tradeoffs

Manually switching dashboards

A developer gets the issue status from the Issues tab, then clicks over to the Pipelines tab, and finally has to copy the MR ID to check the code in the Code tab. This takes 5-10 minutes of context switching.

→ Instead, ask your agent to use list_project_issues, list_project_pipelines, and list_merge_requests sequentially. The agent runs all three tools and synthesizes a single, comprehensive status report in the chat.

Assuming full access

A new user assumes they can see all projects and try to run a global search before verifying their permissions. The search fails, leaving the user unsure if the failure was due to permissions or a bad query.

→ Start by calling get_my_gitlab_profile to confirm your access level. Then, use list_visible_groups and list_visible_projects to scope your search correctly before running search_gitlab_global.

Ignoring the pipeline state

A developer sees a merge request is ready, but doesn't check the build status. They assume the code passed CI/CD, leading to potential integration failures later.

→ Before merging, ask the agent to run list_project_pipelines on the target project. This confirms the most recent build passed successfully, ensuring the code is deployable.

When It Fits, When It Doesn't

Use this if your primary job involves correlating data points across the full development lifecycle—i.e., you need to know if an Issue is linked to a Merge Request, if that MR is associated with a passing Pipeline, and what files were changed. You must be dealing with multi-stage, interconnected project data. Don't use this if you just need to check a single piece of data (e.g., 'What is the status of issue #123?'). For that, a simpler, single-purpose issue tracker or ticketing system is sufficient. If you only need to read file contents, the get_repository_file tool handles that job fine, but this server gives you the whole picture.

Common Questions About GitLab MCP

How do I get a Personal Access Token (PAT) for GitLab? +

Does this support self-hosted GitLab instances? +

Yes! Simply provide your instance's root URL (e.g., https://gitlab.my-company.com) in the 'GitLab Base URL' credential field.

What is a 'Project ID' in GitLab? +

A Project ID is a unique numerical identifier for your project. You can also use the URL-encoded path (e.g., 'group%2Fproject') in most tools.

Can I see pipeline failures via the agent? +

Yes! Use the 'list_project_pipelines' tool to retrieve the status of all recent pipelines, allowing you to quickly identify failed builds.

How do I use the `list_visible_projects` tool to find a project? +

The list_visible_projects tool returns a list of all projects the authenticated user can access. You can then use the project name or ID to feed into other tools, like get_project_details, for deeper analysis.

What happens if I try to read a file with `get_repository_file` but the file doesn't exist? +

If the file doesn't exist, the tool returns a specific error message detailing the missing path. This means your agent client can handle the failure gracefully and prompt you to check the file path.

Does the `search_gitlab_global` tool search across all user types or just code? +

The global search tool indexes projects, issues, and user metadata across your entire GitLab instance. It finds artifacts by keyword, not just code contents.

How does `create_project_issue` handle required fields for a new issue? +

The tool requires specific fields like the project ID, title, and description to create an issue. If any mandatory data is missing, the function call fails, preventing incomplete records.

View all recipes →