Gitlab MCP Server 6tb

🚀 GitLab-MCP-Server

This is a Model Context Protocol (MCP) server that provides integration capabilities with GitLab. It retrieves pipeline failure information and review comments from specific GitLab projects and provides them to AI assistants.

🚀 Quick Start

This MCP server utilizes the GitLab API to provide the following information to AI assistants:

1. Console output of failed jobs in GitLab pipelines.
2. Unresolved review comments on GitLab merge requests (MRs).
3. Changes in GitLab MRs (differences from the current state of the local repository).

By using the MCP functionality, AI assistants can directly access GitLab information and provide more accurate support.

📦 Installation

# Install uv
$ curl -LsSf https://astral.sh/uv/install.sh | sh

$ cd /path/to/this-mcp-server
# Install libraries
$ uv sync

✨ Prerequisites

You need a GitLab access token. You can issue the access token in GitLab settings → Access Tokens. When issuing the token, please check read_api.

✨ Features

1. Retrieve and Fix Pipeline Failure Information (get_pipeline_failed_jobs)

Retrieve the console output of failed jobs in GitLab pipelines. The AI assistant will make corrections based on the retrieved information.

Output:

• Console output of failed jobs (including job names, statuses, and detailed logs).

2. Retrieve and Address Review Comments on MRs (get_review_comments)

Retrieve and address unresolved review comments on GitLab MRs. Resolved comments and comments not associated with files are excluded.

Output:

• Unresolved review comments associated with files on MRs (including commenters, timestamps, comment content, and file location information).

3. Retrieve and Review MR Changes (get_review_changes)

Retrieve the differences from the base commit (base_sha) of a GitLab MR to the current state of the local repository. You can get the differences from the latest local state (including uncommitted changes in progress), not the remote differences. Reviews are conducted based on the retrieved differences.

Output:

• Changes from the base commit of an MR to the current local state (change types and differences for each file).

🔗 Integration with AI Assistants

AI assistants (such as Claude) can call the following functions on this MCP server:

• get_pipeline_failed_jobs(): Retrieve pipeline failure information.
• get_review_comments(): Retrieve review comments on MRs.
• get_review_changes(): Retrieve changes in MRs.

These functions automatically retrieve information about MRs related to the current branch.

⚙️ Configuration in Claude for Desktop

Add the following configuration to claude_desktop_config.json:

{
    "mcpServers": {
        "gitlab-mcp": {
            "command": "uv",
            "args": [
                "--directory",
                "/path/to/this-mcp-server",
                "run",
                "main.py"
            ],
            "env": {
                "GITLAB_URL": "your_gitlab_url",
                "GITLAB_PROJECT_NAME": "gitlab_project_name",
                "GITLAB_API_KEY": "your_gitlab_api_key",
                "GIT_REPO_PATH": "/path/to/git/repo"
            }
        }
    }
}

⚠️ Important Note

In the above configuration example, please replace the following values appropriately:

• your_gitlab_api_key: GitLab API access token.
• /path/to/git/repo: Absolute path to the local Git repository.
• /path/to/this-mcp-server: Absolute path to the directory of this MCP server.