CERN GitLab MCP Server
An MCP server that connects LLMs to CERN GitLab to discover and analyze High Energy Physics code, documentation, and analysis examples. It provides 14 tools for repository browsing, dependency parsing, and CI/CD configuration analysis.
README
<p align="center"> <img src="https://raw.githubusercontent.com/MohamedElashri/cerngitlab-mcp/main/icons/icon.svg" alt="CERN GitLab MCP Server" width="160" /> </p>
<h1 align="center">CERN GitLab MCP Server</h1>
<p align="center"> An <a href="https://modelcontextprotocol.io/">MCP</a> server that connects LLMs to <a href="https://gitlab.cern.ch">CERN GitLab</a> for discovering HEP code, documentation, and analysis examples. </p>
Features
- 14 MCP tools for searching, browsing, and analyzing CERN GitLab repositories
- Public access — works without authentication for public repositories
- HEP-focused — dependency parsing for Python and C++ ecosystems, binary detection for
.rootfiles - Robust — rate limiting, retries with exponential backoff, graceful error handling
Installation
Requires Python 3.10+.
Quickstart (recommended)
No installation needed — just use uvx to run directly:
uvx cerngitlab-mcp
From PyPI
pip install cerngitlab-mcp
From source
git clone https://github.com/MohamedElashri/cerngitlab-mcp
cd cerngitlab-mcp
uv sync
Configuration
All settings are configured via environment variables prefixed with CERNGITLAB_:
| Variable | Default | Description |
|---|---|---|
CERNGITLAB_GITLAB_URL |
https://gitlab.cern.ch |
GitLab instance URL |
CERNGITLAB_TOKEN |
(empty) | Personal access token (optional) |
CERNGITLAB_TIMEOUT |
30 |
HTTP timeout in seconds |
CERNGITLAB_MAX_RETRIES |
3 |
Max retries for failed requests |
CERNGITLAB_RATE_LIMIT_PER_MINUTE |
300 |
API rate limit |
CERNGITLAB_LOG_LEVEL |
INFO |
Logging level |
Authentication
The server works in two modes:
- Without token — Access to all public repositories. Sufficient for most HEP code discovery.
- With token — Additional access to internal/private projects, code search, and wiki pages.
To create a token:
- Go to https://gitlab.cern.ch/-/user_settings/personal_access_tokens
- Create a token with
read_apiscope - Set
CERNGITLAB_TOKEN=glpat-xxxxxxxxxxxx
Note: The code search (
search_code) and wiki (get_wiki_pages) tools require authentication on CERN GitLab.
Usage
Claude Desktop
Add to your claude_desktop_config.json:
{
"mcpServers": {
"cerngitlab": {
"command": "uvx",
"args": ["cerngitlab-mcp"],
"env": {
"CERNGITLAB_TOKEN": "glpat-xxxxxxxxxxxx"
}
}
}
}
Claude Code
claude mcp add cerngitlab-mcp -- uvx cerngitlab-mcp
To include authentication:
claude mcp add cerngitlab-mcp -e CERNGITLAB_TOKEN=glpat-xxxxxxxxxxxx -- uvx cerngitlab-mcp
GitHub Copilot
Add to your VS Code settings.json:
{
"mcp": {
"servers": {
"cerngitlab": {
"command": "uvx",
"args": ["cerngitlab-mcp"],
"env": {
"CERNGITLAB_TOKEN": "glpat-xxxxxxxxxxxx"
}
}
}
}
}
Or add a .vscode/mcp.json to your project:
{
"servers": {
"cerngitlab": {
"command": "uvx",
"args": ["cerngitlab-mcp"],
"env": {
"CERNGITLAB_TOKEN": "glpat-xxxxxxxxxxxx"
}
}
}
}
Gemini CLI
Add to your ~/.gemini/settings.json:
{
"mcpServers": {
"cerngitlab": {
"command": "uvx",
"args": ["cerngitlab-mcp"],
"env": {
"CERNGITLAB_TOKEN": "glpat-xxxxxxxxxxxx"
}
}
}
}
Direct usage
# Run with uvx (no install needed)
uvx cerngitlab-mcp
# Or if installed from PyPI
cerngitlab-mcp
# Or from source
uv run cerngitlab-mcp
# With authentication
CERNGITLAB_TOKEN=glpat-xxx uvx cerngitlab-mcp
Tools
| Tool | Description | Auth required |
|---|---|---|
search_repositories |
Search public repositories by keywords, topics, or language | No |
get_repository_info |
Get repository details (languages, stats, license) | No |
list_repository_files |
Browse the file tree of a repository | No |
get_file_content |
Retrieve file content with binary detection | No |
get_repository_readme |
Get the README file (tries common filenames) | No |
search_code |
Search for code across repositories | Yes* |
get_wiki_pages |
Access repository wiki pages | Yes |
analyze_dependencies |
Parse dependency files (Python, C++, Fortran) | No |
get_ci_config |
Retrieve and analyze .gitlab-ci.yml |
No |
get_build_config |
Find build config files (CMake, Make, setuptools, etc.) | No |
list_releases |
List releases from a repository | No |
get_release |
Get details of a specific release | No |
list_tags |
List repository tags with optional filtering | No |
test_connectivity |
Test connection to the GitLab instance | No |
* search_code works without auth for project-scoped search (uses fallback grep). Global search requires authentication.
For detailed parameter documentation, see dev.md.
Example Prompts
Search for repositories
"Search CERN GitLab for Python repositories related to ROOT analysis and show me the most starred ones"
Understand a project
"Get the README and file structure of the lhcb/DaVinci project on CERN GitLab"
Find fitting examples
"Search for repositories on CERN GitLab that use RooFit and show me example fitting code"
Check dependencies
"What are the dependencies of the lhcb/allen project? Show me the CMakeLists.txt and any Python requirements"
Track releases
"List the recent releases of lhcb/DaVinci and show me the release notes for the latest version"
Explore CI/CD
"Get the CI/CD configuration of the lhcb/allen project and explain the pipeline stages"
Find framework configurations
"Search for Gaudi framework configuration files on CERN GitLab and show me examples"
Development
See dev.md for development setup, project structure, testing, and release instructions.
License
AGPL-3.0
推荐服务器
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。
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。
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。
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。
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。
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。
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。
e2b-mcp-server
使用 MCP 通过 e2b 运行代码。
Neon MCP Server
用于与 Neon 管理 API 和数据库交互的 MCP 服务器
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。