MCP Elicitations Demo Server
A demonstration server that showcases how to collect user input dynamically using the Model Context Protocol (MCP) elicitation system across tools, resources, and prompts.
README
MCP Everything Server
A comprehensive reference implementation for the Model Context Protocol (MCP) that demonstrates all features of the protocol. This server showcases the versatility and extensibility of MCP, demonstrating how it can be used to give Large Language Models (LLMs) secure, controlled access to tools and data sources.
This server is built with the TypeScript MCP SDK.
Features
This server implements a comprehensive set of MCP features including:
- Tools: Various tools including echo, add, long-running operations, environment variable printing, LLM sampling, and more
- Prompts: Simple and complex prompts with arguments, resource references, and embedded content
- Resources: Static resources with pagination support, resource templates, and subscriptions
- Logging: Configurable logging levels with real-time notifications
- Completions: Auto-completion for prompt arguments and resource references
- Elicitation: Interactive user input collection with structured schemas
- Annotations: Content annotations for priority and audience targeting
- Progress Tracking: Progress notifications for long-running operations
- Real-time Updates: Resource update notifications and stderr streaming
Installation
npm install -g @modelcontextprotocol/server-everything
Usage
Standard I/O
mcp-server-everything
Server-Sent Events (SSE)
npm run start:sse
HTTP with Streaming Support
npm run start:streamableHttp
Development
- Clone the repository:
git clone https://github.com/modelcontextprotocol/servers.git
cd servers
- Install dependencies:
npm install
- Build the server:
npm run build
- Run the server:
npm start
Configuration
The server supports various configuration options and can be used to test and demonstrate MCP capabilities. It includes:
- 100 static resources (mix of text and binary)
- Multiple tool implementations
- Various prompt templates
- Real-time notifications and updates
- Progress tracking for long operations
Docker
Build and run with Docker:
docker build -t mcp-everything-server .
docker run -p 3000:3000 mcp-everything-server
License
MIT - See LICENSE file for details.
Contributing
See CONTRIBUTING.md for contribution guidelines.
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。