ibm-cloud-docs-mcp-server
Enables AI clients to search and retrieve IBM Cloud documentation, converting pages to clean Markdown.
README
IBM Cloud Documentation MCP Server
A Model Context Protocol (MCP) server that provides AI clients with the ability to search and retrieve IBM Cloud documentation. This server exposes IBM Cloud's documentation as standardized tools that can be used by LLMs and AI applications.
Features
- Search Documentation: Search IBM Cloud's extensive documentation library with pagination support
- Read Documentation: Retrieve and convert documentation pages to clean Markdown format
- Clean API: Simplified responses containing only essential information
- MCP Compatible: Works with any MCP-compatible client (Claude Desktop, VSCode with MCP extension, etc.)
Installation
bun install
Usage
Running the MCP Server
bun run mcp
This starts the MCP server using stdio transport, making it compatible with MCP clients.
Running the Demo Application
bun run index.ts
This runs a standalone demo that searches for Ubuntu VSI documentation and converts it to Markdown.
Testing the MCP Server
bun run test-mcp
This runs a test client that connects to the MCP server and demonstrates both tools in action.
MCP Tools
search_documentation
Search IBM Cloud documentation for relevant topics and articles.
Parameters:
query(string): Search query for IBM Cloud documentationlimit(number, optional): Maximum number of results to return (1-20, default 10)offset(number, optional): Number of results to skip for pagination (default 0)
Returns: JSON object with search results, total count, and pagination info.
read_documentation
Read and convert IBM Cloud documentation page to clean Markdown format.
Parameters:
href(string): Documentation href from search results (e.g., '/docs/containers?topic=containers-getting-started')
Returns: Clean Markdown content of the documentation page.
Integration with Claude Desktop
To use this MCP server with Claude Desktop, add the following to your claude_desktop_config.json:
{
"mcpServers": {
"ibm-cloud-docs": {
"command": "bun",
"args": ["--cwd", "/absolute/path/to/this/project", "run", "mcp"]
}
}
}
Example Usage
-
Search for documentation:
Search for "kubernetes deployment" with limit 5 -
Read specific documentation:
Read documentation from href "/docs/containers?topic=containers-getting-started" -
Paginated search:
Search for "virtual servers" with offset 10 to get the next page
Architecture
- mcp-server.ts: Main MCP server implementation
- ibm-cloud-api.ts: API client for IBM Cloud documentation
- utils.ts: Utility functions for URL parsing and HTML-to-Markdown conversion
- types.ts: TypeScript type definitions
The server reuses robust API integration code while providing a clean, simplified interface through the MCP protocol.
Requirements
- Bun runtime
- Node.js 18+ (for MCP SDK compatibility)
- Internet connection (for accessing IBM Cloud documentation APIs)
License
This project was created using bun init in bun v1.2.15.
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。