Versions:
GitLab MCP Server, published by José M. Requena Plens, is a Model Context Protocol server that exposes the GitLab REST API v4 and GraphQL APIs as tools for AI assistants. Its purpose is to connect assistants such as Claude, Cursor, VS Code with Copilot, and any other MCP client to GitLab.com or self-managed GitLab instances, enabling those clients to interact with GitLab programmatically through the standardized MCP interface. The server covers the full REST API v4 and GraphQL surface with automatic edition gating, providing roughly 850 to 1080 tools depending on the GitLab tier in use. In addition to tools, it offers 45 MCP resources, 37 prompts, argument completions, progress notifications, and resource subscriptions, giving AI clients a broad and structured set of capabilities for querying and working with GitLab data. The software is distributed as a single static binary and supports both stdio and HTTP transports, which allows it to fit into different deployment and integration scenarios. For safer operation, it includes read-only and safe (mutation-preview) modes, letting administrators and users control how AI assistants may modify data. Per-client configuration examples are provided in the documentation to simplify setup across supported assistants and MCP clients. The current version is 3.0.0, and a total of 14 versions have been released, indicating ongoing development and iteration of the project. Within a software catalog, GitLab MCP Server falls under the category of AI integration and developer tooling, specifically MCP servers that bridge large language model assistants with external platforms. Typical use cases include allowing an AI assistant to inspect repositories, issues, merge requests, pipelines, and other GitLab entities; automating workflows through assistant-driven API calls; previewing mutations before applying them; and providing consistent, protocol-based access to GitLab functionality across multiple AI clients and environments.
Tags: