gitlab-mcp-server

gitlab-mcp-server

Enables interacting with GitLab repositories, merge requests, and code through natural language using MCP. Supports authentication with personal access tokens or OAuth2, and provides tools for listing projects, reading repository code, and analyzing merge request lifetimes.

Category
访问服务器

README

mcp-gitlab-server

⚠️ Development Stage Notice
This project is currently in development stage. Features and APIs may change without notice. Use with caution in production environments.

GitLab MCP server based on python-gitlab.

Install

Using Personal Access Token (Most Common)

{
  "mcpServers": {
    "GitLab": {
      "command": "uvx",
      "args": [
        "gitlab-mcp-server"
      ],
      "env": {
        "GITLAB_TOKEN": "<your GitLab personal access token>",
        "GITLAB_URL": "https://gitlab.com"
      }
    }
  }
}

Using OAuth2 Token

{
  "mcpServers": {
    "GitLab": {
      "command": "uvx",
      "args": [
        "gitlab-mcp-server"
      ],
      "env": {
        "GITLAB_OAUTH_TOKEN": "<your GitLab OAuth2 token>",
        "GITLAB_URL": "https://gitlab.com"
      }
    }
  }
}

For self-hosted GitLab instances, set GITLAB_URL to your GitLab instance URL (e.g., https://gitlab.example.com). If not set, it defaults to https://gitlab.com.

Authentication

This MCP server supports two authentication methods:

Method 1: Personal Access Token (Recommended for most users)

  1. Create a GitLab Personal Access Token:

    • Go to GitLab → User Settings → Access Tokens
    • Create a token with read_api scope (minimum required)
  2. Set the GITLAB_TOKEN environment variable to your token value

Method 2: OAuth2 Token (For OAuth2 applications)

  1. If you have an OAuth2 token from a GitLab OAuth2 application flow
  2. Set the GITLAB_OAUTH_TOKEN environment variable to your OAuth2 token value

To create an OAuth2 application:

  1. Go to GitLab → User Settings → Applications
  2. Create a new application with appropriate scopes (read_api minimum)
  3. Use the OAuth2 flow to obtain an access token
  4. Use that token as GITLAB_OAUTH_TOKEN

Environment Variables

  • GITLAB_TOKEN - Your GitLab Personal Access Token (if using personal token auth)
  • GITLAB_OAUTH_TOKEN - Your GitLab OAuth2 Token (if using OAuth2 auth)
  • GITLAB_URL - GitLab instance URL (defaults to https://gitlab.com)

Note: The server will first check for GITLAB_OAUTH_TOKEN, and if not found, will use GITLAB_TOKEN. You only need to set one of these.

Tools

Repository & Project Management

  • list_projects - List GitLab projects accessible to the authenticated user (supports filtering by owned/starred and pagination)
  • list_groups - List GitLab groups accessible to the authenticated user
  • list_group_projects - List all projects within a specific GitLab group
  • get_user_info - Get information about the authenticated user
  • search_repositories - Search for GitLab repositories by name, description, or keywords
  • get_repository_details - Get detailed information about a specific repository

Code Access

  • read_repository_code - Read the complete code structure and content of a repository with filtering options
  • read_repository_file - Read the content of a specific file from a repository

Merge Request Analytics

  • list_merge_requests - List merge requests for a repository with filtering by state, ordering, and pagination
  • get_merge_request_analytics - Calculate comprehensive merge request lifetime statistics including average time from creation to merge

Repository Code Reading Features

The MCP server provides powerful code reading capabilities:

read_repository_code

Reads the entire repository structure and file contents with advanced filtering:

  • Selective Reading: Include/exclude files using glob patterns (e.g., *.py,*.js or exclude *.log,node_modules/*)
  • Size Limits: Control maximum files to read and file size limits to prevent overwhelming responses
  • Branch/Tag Support: Read from any branch, tag, or commit SHA
  • Directory Filtering: Focus on specific directories within the repository
  • Smart Filtering: Automatically excludes common build artifacts, logs, and binary files

Example Usage:

  • Read only Python files: include_patterns: "*.py"
  • Exclude tests: exclude_patterns: "*test*,*spec*"
  • Read specific directory: path: "src/"
  • Different branch: ref: "develop"

read_repository_file

Reads individual files from a repository:

  • Single File Access: Get content of specific files quickly
  • Metadata Included: File size, encoding, last commit info
  • Binary Detection: Safely handles binary files
  • Flexible References: Works with branches, tags, or commit SHAs

Limits & Safety:

  • Default: Max 50 files, 100KB per file
  • Configurable limits to prevent timeouts
  • Automatic binary file detection
  • Graceful error handling for large repositories

Merge Request Analytics Features

The MCP server provides comprehensive merge request analytics capabilities:

list_merge_requests

Lists merge requests with advanced filtering and sorting:

  • State Filtering: Filter by state (opened, closed, merged, all)
  • Flexible Sorting: Order by creation date, update date, or title
  • Rich Metadata: Includes author, assignees, reviewers, labels, and voting info
  • Pagination Support: Control number of results returned

get_merge_request_analytics

Calculates detailed merge request lifetime statistics:

  • Lifetime Analysis: Average, median, min/max time from creation to merge
  • Statistical Distribution: 25th, 75th, and 90th percentile analysis
  • Time Period Control: Analyze MRs from the last N days (default: 90 days)
  • Detailed Breakdown: Individual MR details with exact lifetime calculations
  • Multiple Time Units: Results provided in both hours and days

Example Questions You Can Answer:

  • "What is the average lifetime of an MR in the acapulco repository?"
  • "How long do merge requests typically take to get merged?"
  • "What's the distribution of merge request lifetimes?"
  • "Which merge requests took the longest to merge?"

Analytics Output:

  • Average, median, min, max merge times
  • Percentile analysis (25th, 75th, 90th)
  • Sample merge requests with individual lifetimes
  • Complete dataset for further analysis

Development

# Clone the repository
git clone <repo-url>
cd gitlab-mcp-server

# Install dependencies
uv sync

# Run in development mode
uv run python -m mcp_gitlab_server

Testing Configuration

For development and testing purposes, you can use a local wheel file installation:

MCP Configuration for Testing:

{
  "mcpServers": {
    "GitLab": {
      "command": "uvx",
      "args": [
        "--from", "/path/to/your/gitlab-mcp-server/dist/mcp_gitlab_server-0.1.0-py3-none-any.whl",
        "gitlab-mcp-server"
      ],
      "env": {
        "GITLAB_TOKEN": "your-gitlab-token-here",
        "GITLAB_URL": "https://gitlab.com"
      }
    }
  }
}

Build and Test Steps:

# Build the wheel file
uv build

# Test with MCP client (like Claude Desktop)
# Update your MCP configuration with the local wheel path
# The wheel file will be in ./dist/ directory

# For quick testing, verify the server starts:
uv run python -m mcp_gitlab_server

推荐服务器

Baidu Map

Baidu Map

百度地图核心API现已全面兼容MCP协议,是国内首家兼容MCP协议的地图服务商。

官方
精选
JavaScript
Playwright MCP Server

Playwright MCP Server

一个模型上下文协议服务器,它使大型语言模型能够通过结构化的可访问性快照与网页进行交互,而无需视觉模型或屏幕截图。

官方
精选
TypeScript
Magic Component Platform (MCP)

Magic Component Platform (MCP)

一个由人工智能驱动的工具,可以从自然语言描述生成现代化的用户界面组件,并与流行的集成开发环境(IDE)集成,从而简化用户界面开发流程。

官方
精选
本地
TypeScript
Audiense Insights MCP Server

Audiense Insights MCP Server

通过模型上下文协议启用与 Audiense Insights 账户的交互,从而促进营销洞察和受众数据的提取和分析,包括人口统计信息、行为和影响者互动。

官方
精选
本地
TypeScript
VeyraX

VeyraX

一个单一的 MCP 工具,连接你所有喜爱的工具:Gmail、日历以及其他 40 多个工具。

官方
精选
本地
graphlit-mcp-server

graphlit-mcp-server

模型上下文协议 (MCP) 服务器实现了 MCP 客户端与 Graphlit 服务之间的集成。 除了网络爬取之外,还可以将任何内容(从 Slack 到 Gmail 再到播客订阅源)导入到 Graphlit 项目中,然后从 MCP 客户端检索相关内容。

官方
精选
TypeScript
Kagi MCP Server

Kagi MCP Server

一个 MCP 服务器,集成了 Kagi 搜索功能和 Claude AI,使 Claude 能够在回答需要最新信息的问题时执行实时网络搜索。

官方
精选
Python
e2b-mcp-server

e2b-mcp-server

使用 MCP 通过 e2b 运行代码。

官方
精选
Neon MCP Server

Neon MCP Server

用于与 Neon 管理 API 和数据库交互的 MCP 服务器

官方
精选
Exa MCP Server

Exa MCP Server

模型上下文协议(MCP)服务器允许像 Claude 这样的 AI 助手使用 Exa AI 搜索 API 进行网络搜索。这种设置允许 AI 模型以安全和受控的方式获取实时的网络信息。

官方
精选