The Gitlab MCP Server, developed by Jmrplens for the mcp platform, is a free mcps program. Built with Go, it comes as a single static binary or container. This server connects AI clients to GitLab instances, exposing a vast array of GitLab operations as AI-accessible tools. It bridges AI assistants, developers, and DevOps engineers with GitLab workflows, enabling automated interaction with REST v4 and GraphQL APIs. It supports both GitLab.com and self-hosted instances.

What does Gitlab MCP Server connect to?

The Gitlab MCP Server connects directly to GitLab instances, encompassing both GitLab.com and self-hosted GitLab deployments. It functions as an MCP (Model Context Protocol) server, translating GitLab's native APIs into a format accessible by AI. Specifically, it exposes the GitLab REST v4 and GraphQL APIs as MCP tools. This allows AI clients to interact with various GitLab functions, facilitating automated workflows and data retrieval across different GitLab environments.

Core features and tools

The server exposes over 1,000 GitLab operations as AI-accessible tools, covering the full REST and GraphQL API. Specific callable tools include create_issue, get_issue, create_merge_request, get_merge_request, get_merge_request_commits, get_merge_request_diffs, get_merge_request_pipelines, get_pipeline_jobs, manage_pipeline, create_workitem_note, get_workitem_notes, and semantic_code_search. It also enables operations such as drafting releases and triaging pipelines, providing a broad interface for AI-driven automation within GitLab.

Who is it for?

The Gitlab MCP Server is for AI assistants, developers, and DevOps engineers who aim to integrate AI capabilities into their GitLab workflows. It serves those looking to automate tasks related to code management, issue tracking, and CI/CD pipelines through AI. Confirmed compatible AI clients include Claude, Cursor, VS Code, Claude Desktop, Claude Code, Gemini Code Assist, Gemini CLI, GitHub Copilot in VS Code, Kiro IDE and CLI, OpenAI Codex, and Zed. Skip this if you prefer to manage GitLab operations manually without AI integration.

How to get started

Access is granted through a standard OAuth authorization flow. It also supports Personal Access Tokens and uses stdio or HTTP transport protocols.

Use cases

• Code Review and Merge Request Management: AI assistants can review merge requests, check their safety for merging, view status, discussions, commits, and diffs, and create new merge requests.
• Issue Tracking and Management: AI can list open issues, create new issues with detailed information (title, description, assignees, labels, milestone, epic), retrieve detailed issue information, and add comments or internal notes to work items.
• CI/CD Pipeline Monitoring and Triage: AI can check the status of pipelines, list, create, update, retry, cancel, and delete pipelines and jobs, and retrieve job logs and build results.

Gitlab MCP Server empowers AI clients to interact directly with GitLab, automating complex development and operations tasks. Its support for OAuth 2.0 and PATs, coupled with broad API exposure, positions it as a key integration point for AI-driven workflows across various compatible clients. This server offers a direct path for AI assistants to manage code, issues, and pipelines. Try creating an issue or reviewing a merge request using an AI assistant.