What is an MCP Server?

An MCP (Model Context Protocol) server acts as a bridge, allowing AI models to access and interact with external data sources and tools. It standardizes how applications provide context to Large Language Models (LLMs).

How does the Ref MCP Server help AI agents?

The Ref MCP Server provides AI coding tools and agents with access to a comprehensive and up-to-date repository of documentation for APIs, services, libraries, and more. This ensures agents have the most current information for accurate and efficient operation.

What are the key features of the Ref MCP Server?

The key features include ref_search_documentation for searching technical documentation, ref_read_url for reading the full content of web pages, and ref_search_web (optional) as a fallback web search tool.

How do I set up the Ref MCP Server?

You can set it up either via the streamable-http server (experimental) or the local stdio server, with streamable-http being the recommended approach as of April 2025.

What is UBOS and how does Ref MCP Server integrate with it?

UBOS is a full-stack AI Agent Development Platform that helps you orchestrate AI Agents, connect them with your enterprise data, build custom AI Agents with your LLM model, and create Multi-Agent Systems. The Ref MCP Server integrates seamlessly with UBOS to enhance AI agent development workflows.

Can I disable the ref_search_web tool?

Yes, you can disable it by setting the DISABLE_SEARCH_WEB environment variable to true or by setting the disable_search_web=false URL parameter in the streamable-http server.

Where can I find more documentation on the MCP Inspector tool?

Visit the Inspector documentation for detailed setup instructions.

What license is the Ref MCP Server released under?

The Ref MCP Server is released under the MIT license.