Harbor MCP Server
Enables AI models to interact with Harbor container registries to manage projects, repositories, and system configurations in real time. It provides a suite of tools for monitoring health, retrieving statistics, and performing searches across registry data.
README
<h1 align="center" style="font-weight: bold;">Harbor MCP Server (Experimental 🧪)</h1>
This is an experimental, community-driven MCP server for Harbor. It is not an official Harbor project.
We welcome and rely on community contributions—join us to improve and expand this tool!
What is this?
A Model Context Protocol (MCP) server for Harbor, enabling AI models and tools to interact with a Harbor container registry in real time.
Project Links
Quick Start
-
Clone & Install:
git clone https://github.com/bupd/harbor-mcp-server.git cd harbor-mcp-server bun install bun run build -
Set Environment Variables:
- export HARBOR_API_BASE=YOUR_HARBOR_URL (e.g.
https://demo.goharbor.io/api/v2.0) - export HARBOR_AUTH_USER=YOUR_USERNAME
- export HARBOR_AUTH_PASS=YOUR_PASSWORD (or token or secret or whatever you call it)
- export HARBOR_API_BASE=YOUR_HARBOR_URL (e.g.
-
Run:
node build/index.js -
Integrate with your MCP client (e.g. Claude Desktop, Cursor) using the appropriate command and environment variables.
{
"mcpServers": {
"harbor": {
"command": "npx",
"args": [
"-y",
"harbor-mcp"
],
"env": {
"HARBOR_API_BASE": "HARBOR_URL",
"HARBOR_AUTH_USER": "HARBOR_USERNAME",
"HARBOR_AUTH_PASS": "HARBOR_PASSWORD_OR_SECRET"
}
}
}
}
If you have it locally use the below
{
"mcpServers": {
"harbor": {
"command": "node",
"args": [
"YOUR_ABSOLUTE_PATH/harbor-mcp-server/build/index.js"
],
"env": {
"HARBOR_API_BASE": "HARBOR_URL",
"HARBOR_AUTH_USER": "HARBOR_USERNAME",
"HARBOR_AUTH_PASS": "HARBOR_PASSWORD_OR_SECRET"
}
}
}
}
Run with Docker
"mcpServers": {
"harbor-mcp": {
"command": "docker",
"args": [
"run",
"-i",
"--rm",
"-e", "HARBOR_API_BASE",
"-e", "HARBOR_AUTH_USER",
"-e", "HARBOR_AUTH_PASS",
"bupd/harbor-mcp-server:latest"
],
"env": {
"HARBOR_API_BASE": "YOUR_HARBOR_URL",
"HARBOR_AUTH_USER": "YOUR_HARBOR_USERNAME",
"HARBOR_AUTH_PASS": "YOUR_HARBOR_PASSWORD_OR_SECRET"
}
}
}
Features / Commands
The following Harbor operations are available as MCP tools:
get-health: Check Harbor instance health.get-statistics: Get project and repository statistics.list-projects: List projects (with filters/pagination).get-project: Get details for a specific project.get-project-summary: Get a summary of a project.list-project-members: List members of a project.list-repositories-in-project: List repositories in a project.list-quotas: List storage quotas for projects.search: Search for projects and repositories by name.get-configurations: Get system configurations (admin only).update-configurations: Update system configurations (admin only, confirmation required).get-volumes: Get system volume info (total/free size).
Contributing
- This project is maintained by the community, for the community.
- We are looking for contributors!
- Open an issue or PR to get involved.
For more details look at (CONTRIBUTING.md)[CONTRIBUTING.md]
Disclaimer
This is not an official Harbor MCP server.
It is experimental and under active development by the community.
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。