AI MCP Service New

Oh-my-rime provides a knowledge base query service based on MCP (Model Context Protocol). With MCP, you can let AI editors (such as Cursor, Windsurf, VS Code, etc.) directly access the oh-my-rime documentation knowledge base, getting accurate configuration tutorials and usage guides during AI conversations.

In short: After configuring MCP, you can ask questions about oh-my-rime directly in your AI editor, and the AI will automatically query our knowledge base to answer you.

What is MCP?

MCP (Model Context Protocol) is an open protocol proposed by Anthropic to standardize the interaction between AI models and external data sources and tools. Think of it as a "plugin system" for AI assistants — through MCP, AI can call external tools to retrieve real-time, accurate information instead of relying solely on training data.

Service Information

Item	Description
Endpoint	https://www.mintimate.cc/mcp
Transport	Streamable HTTP
Protocol Version	2025-03-26
Available Tool	query_oh-my-rime — Semantic search of the oh-my-rime knowledge base

Configuration in AI Editors

Cursor

Create a .cursor/mcp.json file in the project root directory, or add it in Cursor's global settings:

{
  "mcpServers": {
    "oh-my-rime-knowledge": {
      "url": "https://www.mintimate.cc/mcp",
      "transport": "streamable-http"
    }
  }
}

Windsurf

Add the following to Windsurf's MCP configuration:

{
  "mcpServers": {
    "oh-my-rime-knowledge": {
      "serverUrl": "https://www.mintimate.cc/mcp",
      "transport": "streamable-http"
    }
  }
}

VS Code (GitHub Copilot)

Create a .vscode/mcp.json file in the project root directory:

{
  "servers": {
    "oh-my-rime-knowledge": {
      "type": "http",
      "url": "https://www.mintimate.cc/mcp"
    }
  }
}

Claude Desktop

Add the following to the Claude Desktop configuration file (config file path: on macOS ~/Library/Application Support/Claude/claude_desktop_config.json, on Windows %APPDATA%\Claude\claude_desktop_config.json):

{
  "mcpServers": {
    "oh-my-rime-knowledge": {
      "url": "https://www.mintimate.cc/mcp",
      "transport": "streamable-http"
    }
  }
}

Note

The configuration format may vary slightly across different AI editors/clients. Please refer to the official documentation of your tool. The examples above are for reference only — the key point is to set the MCP endpoint https://www.mintimate.cc/mcp in the corresponding MCP settings.

Usage

Once configured, you can directly ask questions about oh-my-rime in your AI conversation, and the AI will automatically invoke the query_oh-my-rime tool to query the knowledge base. For example:

• "How do I install oh-my-rime on macOS?"
• "How to configure fuzzy pinyin in oh-my-rime?"
• "How to enable the XiaoHe double pinyin scheme?"
• "What is configuration override in Rime?"

Tool Parameters

The query_oh-my-rime tool supports the following parameters:

Parameter	Type	Required	Description
query	string	✅	Natural language query, supports both Chinese and English. e.g. How to configure oh-my-rime
keyword	string	❌	Keyword filter, multiple keywords separated by semicolons. e.g. macOS;install;Rime

Note

MCP tool invocations are typically handled automatically by the AI assistant. You just need to ask in natural language — there's no need to manually pass these parameters.

Technical Implementation

This MCP service is deployed on EdgeOne Pages Cloud Functions, leveraging the vector semantic search capabilities of the CNB knowledge base. The Markdown content from the documentation repository is automatically vectorized and indexed, enabling high-precision semantic search.

The overall architecture is as follows: