
Aptly MCP Server
Enables AI assistants to manage Debian package repositories through natural language using the Aptly package management tool. Supports repository creation, package operations, snapshot management, publishing, and mirror synchronization.
README
Aptly MCP Server
A Model Context Protocol (MCP) server for managing Debian package repositories using the Aptly package repository management tool.
Overview
This MCP server provides AI assistants with direct access to Aptly functionality through standardized tools and resources. It enables natural language interaction with Debian package repositories, including creation, management, and publishing operations.
Features
MCP Tools
- aptly_create_repo: Create new local repositories
- aptly_list_repos: List all repositories with filtering options
- aptly_add_packages: Add packages to repositories from uploaded files
- aptly_search_packages: Search packages across repositories
- aptly_create_snapshot: Create snapshots from repositories or other snapshots
- aptly_publish_repo: Publish repositories/snapshots for apt consumption
- aptly_create_mirror: Create mirrors of remote repositories
- aptly_upload_files: Upload package files to the server
- aptly_list_mirrors: List all configured mirrors
- aptly_update_mirror: Update mirrors with latest packages
- aptly_delete_repo: Delete repositories with safety checks
MCP Resources
- repositories: Dynamic listing of all repositories with metadata
- packages: Searchable package databases across repositories
- mirrors: Remote mirror status and configuration
- publications: Published repository information and endpoints
- snapshots: Available snapshots with creation dates and descriptions
Installation
cd aptly-mcp
npm install
npm run build
Configuration
Environment Variables
export APTLY_API_URL=http://localhost:8080 # Default Aptly API endpoint
export APTLY_AUTH_TOKEN=your_token_here # Optional authentication token
Claude Desktop Configuration
Add to your claude_desktop_config.json
:
{
"mcpServers": {
"aptly": {
"command": "node",
"args": ["/path/to/aptly-mcp/build/index.js"],
"env": {
"APTLY_API_URL": "http://localhost:8080"
}
}
}
}
Remote Aptly Server
For remote Aptly servers with authentication:
{
"mcpServers": {
"aptly": {
"command": "node",
"args": ["/path/to/aptly-mcp/build/index.js"],
"env": {
"APTLY_API_URL": "https://aptly.example.com",
"APTLY_AUTH_TOKEN": "your-auth-token"
}
}
}
}
Usage Examples
Once configured with Claude Desktop, you can use natural language commands:
Repository Management
- "Create a new repository called 'production-packages' for production releases"
- "List all repositories and show their package counts"
- "Delete the 'test-repo' repository"
Package Operations
- "Add all .deb files from the 'uploads' directory to the 'main' repository"
- "Search for packages containing 'nginx' across all repositories"
- "Remove the 'old-package' from the 'main' repository"
Snapshot Management
- "Create a snapshot of the 'main' repository called 'v1.0-release'"
- "List all snapshots and their creation dates"
- "Create a merged snapshot from 'main' and 'contrib' repositories"
Publishing
- "Publish the 'main' repository to the 'stable' distribution"
- "Update the published 'testing' repository with the latest snapshot"
- "List all published repositories and their endpoints"
Mirror Management
- "Create a mirror of the Ubuntu main repository"
- "Update all mirrors to get the latest packages"
- "Show the status of the 'ubuntu-main' mirror"
Development
Building
npm run build
Testing
npm run inspector # Launch MCP inspector for testing
Watching for Changes
npm run watch
API Documentation
The server connects to Aptly's REST API. Ensure your Aptly instance is running with API enabled:
aptly api serve -listen=":8080"
Security Considerations
- Use Unix domain sockets instead of network ports in production
- Implement proper authentication for remote Aptly servers
- Consider using a reverse proxy with HTTPS for production deployments
- Validate all inputs to prevent injection attacks
License
MIT License
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。