A Model Context Protocol (MCP) server that lets AI agents interact with GitLab projects, merge-requests and issues (list, diff, comment, edit, etc.).
https://github.com/kopfrechner/gitlab-mr-mcpStop switching between your AI chat and GitLab's web interface. This MCP server connects your AI directly to GitLab projects, letting you manage merge requests and issues without leaving your development flow.
You're deep in code review mode. Jump to GitLab to check the MR diff, back to your AI to discuss the changes, back to GitLab to add comments, back to your AI for suggestions. Rinse and repeat until your focus is completely shattered.
GitLab MR MCP eliminates this workflow friction by putting GitLab directly in your AI's toolkit.
Complete MR Management: Your AI can fetch MR details, grab diffs, add both general and line-specific comments, and even update titles and descriptions. No more copy-pasting code snippets for context.
Smart Code Reviews: Ask your AI to "review the latest MR and suggest improvements to the error handling." It pulls the diff, analyzes the changes, and adds targeted comments directly to the relevant lines.
Issue Triage Automation: "Check all open issues and categorize them by severity" becomes a single command instead of clicking through dozens of issue pages.
Bulk Operations: Need to add consistent feedback across multiple MRs? Your AI can iterate through them programmatically rather than you clicking through each one manually.
The Deep Dive Review: Instead of explaining complex architectural changes to your AI through screenshots and copy-paste, it can directly examine the MR diff and understand the full context of your changes.
Compliance Checking: Set up workflows where your AI automatically reviews MRs for coding standards, security patterns, or documentation requirements, adding specific feedback exactly where needed.
Status Updates: "Generate a summary of all open MRs and their current status" gives you instant project visibility without manually checking each one.
Automated Responses: Your AI can monitor MRs and automatically respond to common questions or add standard checklists based on the type of changes detected.
Works with Claude Desktop out of the box. Install via Smithery for automatic setup:
npx -y @smithery/cli@latest install @kopfrechner/gitlab-mr-mcp --client claude --config '"{\"gitlabMrMcpToken\":\"YOUR_GITLAB_TOKEN\", \"gitlabMrMcpHost\": \"YOUR_GITLAB_HOST\"}"'
Or configure manually if you prefer to control the setup. Either way, you'll be up and running in under 5 minutes.
The server handles authentication through your GitLab personal access token and supports both GitLab.com and self-hosted instances. Project access is automatically scoped to your token permissions, so your AI only sees what you can see.
Perfect for teams that want to streamline their GitLab workflows without changing their existing processes. Your AI becomes an intelligent interface to GitLab rather than a replacement for it.
The line-specific commenting feature is particularly powerful for code reviews - your AI can add contextual feedback directly to the relevant code sections, making reviews more actionable and reducing back-and-forth communication.
Ready to eliminate GitLab context switching from your development workflow? Your AI is about to become much more useful for project management.